Convenient virtual reality exposure self-training for social anxiety: a randomized controlled study

doi:10.21203/rs.3.rs-4660985/v1

Download PDF

Research Article

Convenient virtual reality exposure self-training for social anxiety: a randomized controlled study

https://doi.org/10.21203/rs.3.rs-4660985/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

People with social anxiety (SA) experience emotions that are disproportionate to the objective threat when facing social situations where they might be scrutinized by others. Virtual reality exposure (VRE) therapy has now been explored as a novel and cost-effective mode of psychological treatment, however, less previous randomized controlled trial has examined whether self-help VRE training using convenient appliances can effectively decrease SA symptoms with a long-term effect. Participants with high SA (n = 61) were randomly allocated to either an intervention group (n = 31) or a waiting list control group (n = 30). Mixed ANOVA revealed a significant interaction between TIME and GROUP for SA and negative emotion. Within-group analysis demonstrated that there was a significant reduction in measures from pre-intervention to post-intervention for the intervention group, however, the difference for the control group did not reach statistical significance. Additionally, participants in the intervention group showed significant improvements in SA symptoms, negative emotions and mental health state from pre-intervention to a one-month follow-up. These findings illustrate that self-help VRE training using convenient appliances is an efficacious way to significantly reduce SA and negative emotions, at least for one month.

Virtual reality

Exposure therapy

Convenient VR appliance

Social anxiety

Self-training

The core emotions of social anxiety disorder are "fear" and "anxiety". Patients often feel emotions that are disproportionate to the objective threat when facing social situations, where they might be scrutinized by others. These fears and anxieties are not sporadic.The symptoms are manifested in profound distress and impairment of life functions lasting several months, and patients may even further worry that these symptoms will lead to negative evaluations by others (American Psychiatric Association, 2022; World Health Organization, 2019).The impact of severe social anxiety extends beyond symptom effects, significantly impairing normal functioning across domains such as education, work, and interpersonal relationships (Ruscio et al., 2008; Ye et al., 2018).

Psychological intervention methods for social anxiety disorder include self-disclosure techniques in cognitive therapy, assertive responses techniques in behavioral therapy, and exposure therapy in Cognitive Behavioral Therapy (CBT) (Prochaska & Norcross, 2018). CBT is typically the primary therapy recommended initially (Liebowitz et al., 1999). The key element of exposure therapy addressing its application in treating social anxiety is to allow patients to gradually adapt to an environment with gradually increasing levels of stimulation and improve anxiety symptoms (Safir et al., 2012).

Exposure therapy has proven to be quite effective in treating social anxiety disorder. Some studies even suggest that pure exposure techniques may be superior to cognitive-behavioral therapy that is in collaboration with other techniques (e.g., Horigome et al., 2020). The authenticity of the characters within exposure scenarios and the complexity of the settings can influence the therapy’s efficacy (Lindner et al.,2017; Poeschl & Doering, 2014). However, exposure therapy also has drawbacks, such as patients avoiding treatment due to excessive fear, difficulties in obtaining or imagining stimuli, and challenges in controlling the intensity of the stimuli (Safir et al., 2012). The existing clinical psychological and pharmacological treatment methods require high human resource costs. Professional services are often expensive, prompting the academic community to explore more cost-effective and efficient alternatives, such as group cognitive behavioral therapy (Zaboski et al., 2019) or self-help intervention programs (Chen et al., 2020).

Virtual reality technology uses electronic devices to provide a visualized, realistic scene, making virtual reality exposure therapy a potential option for intervening in social anxiety (Meyerbröker & Emmelkamp, 2010). Firstly, virtual reality exposure therapy addresses the tendency of patients to avoid treatment. Studies have found that comparing to traditional exposure therapy with real-life or imagined scenarios, virtual reality exposure therapy has a lower refusal rate for specific phobias (Garcia-Palacios et al., 2007). Furthermore, children and adolescents may be more motivated to experience the novel virtual reality devices (Kothgassner & Felnhofer, 2021). Secondly, regarding the difficulty of obtaining or imagining stimuli, virtual reality can transcend existing physical and cognitive limits by using visuals and sound to present the stimuli directly in front of the user (Lindner et al., 2017). Finally, various commercial software now enables free customization of parameters likes scenes, audience size, actions, and expressions (Hinojo-Lucena et al., 2020), indicating that the application of graded exposure principles is no longer reliant solely on verbal instruction.

In recent years, a substantial body of review research has supported the outstanding performance of virtual reality exposure therapy. For example, a meta-analytic study by Chesham et al. (2018) had confirmed that the effectiveness of virtual reality technology in intervening social anxiety disorder, and it was particularly noted that there was no difference in effect size between virtual reality exposure and traditional exposure (whether in real or imagined scenarios) in the samples analyzed. Carl et al. (2019) not only estimated a medium to large effect size for virtual reality exposure therapy versus placebo conditions, but also extended their meta-analysis to encompass all anxiety disorders, as well as post-traumatic stress disorder and panic disorder, which reaffirming the same conclusion that virtual reality technology is a good medium for exposure therapy. Furthermore, the systematic review by Kothgassner and Felnhofer (2021) also supported that virtual reality exposure was essentially equivalent in effectiveness to traditional exposure methods.

However, virtual reality devices were once expensive and cumbersome, limiting the willingness of the general public to use them. The complexity of specialized software in the industry had planted obstacles to innovations, especially when it comes to developing applications for therapy. With the rapid advancement of electronic technology, consumer-grade virtual reality products that the general public can afford come into market, making it feasible to transit virtual reality exposure therapy for social anxiety from research labs to more widespread public use (Lindner et al., 2017). In terms of hardware, these consumer-grade products feature a head-mounted portable design that achieves a depth effect visually through different lenses for each eye. Studies have found that combining them with exposure techniques can indeed alleviate anxiety symptoms (Lindner et al., 2021). Regarding software, the development of smartphones has created a new platform: virtual reality exposure training courses that can be operated through apps, which studies have found to be equally applicable to children (Sülter et al., 2022). Overall, advances in both hardware and software technologies have made it possible for individuals to self-intervene with social anxiety.

The present study

Existing research indicates that virtual reality exposure therapy (VRET) is as effective in treating social anxiety disorder as traditional exposure therapy. Moreover, self-administered treatment using portable devices is much less costly than traditional therapist-led interventions, which is expected to benefit those suffering from social anxiety symptoms. However, it is notable that current virtual reality self-intervention experiments primarily present social situations using 3D animated characters (Kim et al., 2017; Kim et al., 2022; Lindner et al., 2021). While this approach has the advantage of allowing researchers to easily manipulate various levels of stimuli, it can lead to unrealistic character depictions. On the one hand, the absence of facial detail can affect the sense of immersion (Poeschl & Doering, 2014); on the other hand, facial details that are close to real but not completely harmonious may even trigger an additional anxiety or fear known as the "Uncanny Valley" effect (Seyama & Nagayama, 2007). Scenarios featuring real people can better control variables that are irrelevant to the experimental purpose. Therefore, this study uses videos featuring real people to match local social scenes.

The present study employed the sample size calculation software G*power 3.1.9.6 to determine the required number of participants. Repeated measures analysis of variance (ANOVA) (parameter settings: effect size f = 0.25, significance level α = 0.05, power β = 0.8, number of groups = 2, number of measurements = 2), resulted in a required sample size of 34 participants.

The method section is reported in accordance with CONSORT (Consolidated standards of reporting trials) guidelines (Schulz, Altman, & Moher, 2010). The trial received ethical approval from the local university Ethics Committee (HR2-0090-2023) and was registered on the ISRCTN website on May 10, 2023. ISRCTN is one of the top three systems for registering psychology experiments (Bradley et al., 2017). The registration number for this study is 14301295, with the last update recorded on May 19, 2023.A CONSORT flow diagram (see Fig. 1) depicts the progression of participants through the three trial phases: enrollment, intervention, and follow-up. Data was collected between April 2023 and September 2023.

INSERT FIGURE 1 ABOUT HERE

The reasons for choosing college students as the targeted sample are as follows: firstly, most colleges and universities have a relatively consistent semester rhythm to avoid interference by too many unrelated factors in a social setting; secondly, sufficient support resources in this campus ensure that both the physical and mental state of participants is guaranteed under experimental conditions; thirdly, training sessions include various kinds of social situations: from learning and interviews to making friends and dating. College students simply have to face different social issues at each phase. An opportunity sample of participants was recruited by posting an invitation poster in the mental health centers of universities or student communities via WeChat, and participants completed the recruitment questionnaire via the QuestionStar WeChat mini-program. In total, 86 participants completed an online screening questionnaire that determined their eligibility for taking part in the trial. This questionnaire collected demographic information and inclusion/exclusion criteria.

Inclusion criteria: a) Current university students in Shanghai: aged 18 or older, with educational backgrounds ranging from undergraduate to doctoral level; b) High levels of self-report scoring SPIN (≥ 31) as measured through the Social Phobia Inventory (SPIN) (Conner et al., 2000); c) High levels of self-report scoring LSAS (≥ 35) as measured through the Liebowitz Social Anxiety Scale (LSAS) (Liebowitz, 1987); d) Willingness to attend intervention sessions.

Exclusion criteria: a) Severe depression > 29 on Beck Depression Inventory-II (BDI-II) (Beck et al., 1996). Given the high comorbidity between social anxiety disorder and depression, completely excluding participants with any level of depressive tendencies is impractical and hinders the applicability of the intervention (Keller, 2003; Wong et al., 2023). Therefore, we follow the initial setup of the Beck et al. (1996) scale and only exclude participants with the highest level of depressive symptoms; b) Diagnosis of other mental disorders besides social anxiety disorder (e.g., bipolar disorder, psychosis, or schizophrenia); c) High risk of suicide or self-harm: Screened using the risk subscale of the CORE-OM questionnaire; d) Current engagement in psychological therapy or taking prescribed psychoactive medication which as changed in < 6 weeks; e) Strong discomfort with VR videos reported by participants during the pre-test.

The study recruited participants in batches, and to ensure random allocation, each batch of participants was initially numbered based on the order of questionnaire completion. The "Select Cases at Random" function in the statistical software SPSS was then used to divide them into the intervention group and the waiting group (parameters: sample size set to "approximately 50% of all cases"). Finally, qualified participants (n = 61) were randomly allocated to either an intervention group (n = 31) or a waiting list control group (n = 30).

Measures

The Social Phobia Inventory (SPIN). This scale consists of 17 items that encompass three aspects: fear, avoidance, and psychological discomfort (Conner, et al., 2000). It employs a five-point rating scale (0 = not at all, 1 = a little bit, 2 = sometimes, 3 = often, 4 = very often), and the scale score is obtained by summing the scores for each item. A higher score on the SPIN scale indicate greater severity of social anxiety. The SPIN scale exhibits strong psychometric properties, including high test-retest reliability (r = 0.89, p < 0.0001) and internal consistency (α = 0.94).

The Liebowitz Social Anxiety Scale (LSAS). This scale, created by Liebowitz in 1987, is utilized to assess social anxiety levels in social situations and performance contexts. The scale comprises a total of 24 items, with each item evaluating the extent of fear and avoidance on a 4-point rating scale. It includes two separate subscales for fear and avoidance. The scale is known for its good reliability, with test-retest reliability coefficients of 0.779, 0.802, and 0.749 for the total score and the two subscales, respectively. The LSAS has been widely adopted in various fields and is particularly prominent in the domain of Virtual Reality Exposure Therapy (VRET), as evidenced by studies such as Chesham et al. (2018) and Ding & Li (2018). Research conducted in China has also found its applicability among adult populations, including college students, as shown by studies conducted by He & Zhang (2004) and Xiao et al. (2007).

The Clinical Outcomes in Routine Evaluation - Outcome Measure (CORE-OM). This scale, created by Evans et al. in 2002, is utilized to assess the effectiveness of psychological therapy and encompasses a broad spectrum of psychosocial dimensions. The scale consists of 34 items, rated on a 5-point scale. It includes four distinct subscales, measuring Subjective Well-being, Problems/Symptoms, Life Functioning, and Risk/Harm. The CORE-OM demonstrates good reliability, with test-retest reliability coefficients exceeding 0.50. It exhibits strong validity, including both convergent and screening validity in Chinese context (Zhang et al., 2020).

The Beck Depression Inventory-II (BDI-II). This scale, developed by Beck et al. in 1996, is a widely used assessment tool for measuring the severity of depression. The scale comprises a total of 21 items, each rated on a 4-point scale. It demonstrates good reliability, with a high internal consistency reliability coefficient (Cronbach’s α) of 0.90. It also exhibits strong validity, including convergent validity, with coefficients exceeding 0.60, discriminant validity, with coefficients exceeding 0.4, and criterion validity.

The Positive and Negative Affect Schedule (PANAS). This scale, developed by Watson et al. in 1988, is a tool used to separately measure positive and negative emotions. The scale consists of a total of 20 items, with each item rated on a 5-point scale. It includes two subscales, one for positive emotions and one for negative emotions. PANAS also exhibits strong validity, including criterion validity. Research conducted in China suggests its applicability among the Chinese adult population (Huang et al., 2003).

The Simulator Sickness Questionnaire (SSQ). This scale, developed by Kennedy in 1993, is designed to measure the degree of physical discomfort experienced after viewing virtual reality imagery. The scale consists of a total of 16 items, with each item rated on a 4-point scale. It exhibits strong validity, including structural validity and criterion validity, as supported by research (Sevinc & Berkman, 2020). Studies conducted in China suggest its applicability among the Chinese adult population, including college students (Zhang et al., 2022).

Physiological Metrics. Cardiovascular reactivity is a common physiological indicator for evaluating social anxiety. Most studies support that higher heart rate and blood pressure variability reflect higher levels of social anxiety (Gramer et al., 2012; Liu et al., 2023; Lu et al., 2022; Kim et al., 2017; Sülter et al., 2022; Yoon & Quartana, 2012), but there are also some different opinions (Gramer, 2006; Gramer & Sprintschnik, 2008). In order to explore cardiovascular reactivity, the study utilized the Xiaomi BPX1 Smart Arm Electronic Blood Pressure Monitor to measure the participants' heart rate and blood pressure. This blood pressure monitor employs the oscillometric method for heart rate measurement and has a blood pressure measurement range from 0 to 300 millimeters of mercury (mmHg).

VR apparatus & Stimuli

The VR equipment used in this study is the Thousand Wonders G04VR headset, which features a head-mounted design and supports the viewing of 360° panoramic VR videos in conjunction with a smartphone. The stimulus materials in the mini-program consist of 360° videos displayed as separate left and right images, with the VR headset presenting each image on the respective left and right lenses of the VR viewer. When users put on the VR headset, the left eye can only see the left image, and the right eye can only see the right image, creating a stereoscopic visual effect in this manner.

The bare VR headset weighs 282g and has dimensions of 190mm×130mm×90mm. The lenses are made of resin aspherical optical lenses with a diameter of 40mm, and the field of view is 108° (the actual field of view may vary based on the size of the smartphone and the tightness of the head strap). The VR headset features diopter adjustment, suitable for individuals with myopia up to -500° who can use it without prescription glasses. The interpupillary distance (IPD) adjustment range is 60–70 mm. This VR headset is compatible with smartphones running various systems, with the optimal experience achieved on smartphones with a display size of 4.76.7 inches.

Based on the research by Ruscio et al. (2008), the prevalence rates for primary types and timings of social anxiety situations are as follows: Meeting new people 16.8%, Conversing with authority Figs. 17.4%, Speaking loudly in meetings or classes 19.5%, Attending parties 13.4%, Public speaking or performing 21.2% and so on. This study excluded scenes that were not suitable for presentation in video format and filmed the following videos (see Fig. 2.):VR Adaptation Test (1 video, No-person scenario);Pretest, Posttest, and Follow-up Tests (Total of 3 videos, all interview scenarios);Exposure Intervention Sessions (Total of 20 videos) (Eye contact: 2, Initial interview: 2, Meeting new friends: 2, Female authority figure: 2, Aunt's probing: 2, Male authority figure: 2, Arguing with uncle: 2, Dating: 4, Impatient leader: 2).

INSERT FIGURE 2 ABOUT HERE

Regarding the virtual reality scene exposure design, Kim et al.'s (2017) study on self-help interventions for social anxiety disorder was divided into the listening phase and the response phase. In the listening phase, characters in the videos presented social situations to trigger fear and anxiety in participants. In the response phase, participants were required to speak according to the given task instructions, while the video characters remained mostly silent. Following this scene exposure design, in each video segment of this study, a listening phase or simple instructions were presented first. Participants' task was to speak during the response phase, while the video characters maintained basic silence or engaged in other distracting actions and negative reactions according to the pre-designed exposure levels (see the script sample in supplementary materials).

Procedure

For the intervention group, the initial pretest was conducted on the first day in a laboratory at East China Normal University. The remaining training sessions were self-administered by participants after they received the VR goggles. On the 14th day, they returned to the laboratory for the posttest and to return the VR goggles. Within one month of completing the posttest, participants received the VR goggles again to complete follow-up exercises and tests on the 43rd and 44th days, respectively.

For the waitlist control group, the initial pretest was conducted on the first day and the posttest on the 14th day, both in a laboratory setting. After the posttest, participants took the self-training sessions as same as the intervention group.

Analysis

Statistical analyses were performed using SPSS Statistics (Version 22) software. Primary analyses were 2*2 Mixed Analysis of Variance (ANOVA) with Group (intervention, waiting control) as the between-group factor and Time (pre-intervention, post-intervention) as the within-subject factor for all outcome variables. For the intervention group (pre-intervention: n = 31; post-intervention: n = 25; follow-up: n = 21), one way ANOVA with Time (pre-intervetion, post-intervention, follow-up) was conducted to analyze the improvement of the intervention process.

Analysis of variance revealed no significant differences between the intervention group and the waiting control group in terms of demographic characteristics or pre-intervention SPIN, LASK, and CORE (see Table 1).

Table 1

Demographic/clinical characteristics of participants per condition at pre-intervention
Demographic/clinical variable	Intervention group (N = 25)	Control group (N = 29)
Gender % Female	22 (88.00%)	25 (86.21%)
Mean Age(SD)	22.32 (3.01)	23.03 (2.87)
Education, n(%)
College	12 (48.00%)	11 (37.93%)
Undergraduate	11 (44.00%)	15 (51.72%)
Postgraduate	2 (8.00%)	3 (10.34%)
CORE-OM(SD)	46.64 (14.39)	47.41 (14.77)
SPIN(SD)	40.08 (10.62)	41.48 (9.25)
LSAS(SD)	70.96 (19.54)	75.52 (22.55)
PANAS-P(SD)	27.12 (4.48)	26.52 (5.57)
PANAS-N(SD)	27.84 (6.67)	28.93 (7.16)
Note: CORE-OM = The Clinical Outcomes in Routine Evaluation - Outcome Measure; SPIN = The Social Phobia Inventory; LSAS = The Liebowitz Social Anxiety Scale; PANAS-P = The Positive and Negative Affect Schedule-Positive; PANAS-N = The Positive and Negative Affect Schedule-Negative.

INSERT Table ABOUT HERE

Analyses of SPIN and LSAS scores revealed that there were statistically significant main effects for GROUP (SPIN: F(1, 52) = 6.528, p = 0.014, η² = 0.112; LSAS: F(1, 52) = 4.133, p = 0.047, η² = 0.074), TIME (SPIN: F(1, 52) = 32.970, p < 0.001, η² = 0.388; LSAS: F(1, 52) = 21.26, p < 0.001, η² = 0.29) and interaction between GROUP and TIME (SPIN: F(1, 52) = 18.182, p < 0.001, η² = 0.259; LSAS: F(1, 52) = 9.678, p < 0.03, η² = 0.157).

The simple effect analysis found that the intervention group had a significant decrease on the SPIN (t(1, 52) = 46.608, p < 0.001,η² = 0.473, M_SPINpre = 40.08, SD = 10.62; M_SPINpost = 28.64, SD = 10.21)and LSAS scores (t(1, 52) = 27.757, p < 0.001, η² = 0.348, M_LSASpre = 70.96 SD = 19.54; M_LSASpost= 55.16, SD = 18.39); however, the control group had no significant improvement on SPIN (t(52) = 1.179, p = 0.282, η² = 0.022; M_SPINpre = 41.48, SD = 9.25; M_SPINpost = 39.79, SD = 9.74) and LSAS scores (t(52) = 1.215, p = 0.275, η² = 0.023; M_LSASpre = 75.52, SD = 22.55; M_LSASpost = 72.45, SD = 22.87).

Analysis of CORE-OM scores revealed a significant TIME effect, F_{(1, 52)} = 11.123, p = 0.002, η² = 0.176. The pre-score of CORE-OM (M_corepre = 46.64, SD = 14.39) was significantly higher than the post-score of CORE-OM (M_corepost = 39.04, SD = 12.49); however, the main effect of GROUP and the interaction between GROUP and TIME did not reached significance (all ps < 0.05).

Analysis of PANAS scores revealed a significant TIME effect (positive emotions: F(1, 52) = 9.896, p = 0.003, η² = 0.160; negative emotions: F(1, 52) = 27.877, p < 0.001, η² = 0.349) and a non-significant GROUP effect (p > 0.05). In details, the pre-score of PANAS-P (M_PANAS−Ppre = 27.12, SD = 4.48) was significantly lower than the post-score of PANAS-P (M_{PANAS−Ppost} = 29.40, SD = 4.84). Meanwhile, the pre-score of PSNAS-N (M_PANAS−Npre = 27.84, SD = 6.67) was significantly higher than the post-score of PSNAS-N (M_{PANAS−Npost} = 21.12, SD = 5.78).

We also found a significant interaction effect of TIME and GROUP on PANAS-N, F_{(1, 52)} = 9.545, p = 0.003, η² = 0.155. The simple effect analysis found that the intervention group had a significant decrease on the PANAS-N scores (M_PANAS−Npre= 27.84, SD = 6.67; M_{PANAS−Npost} = 21.12, SD = 5.78); however, the control group had no significant improvement (M_PANAS−Npre = 28.93, SD = 7.16; M_{PANAS−Npost} = 27.17, SD = 5.28). The interaction effect of TIME and GROUP on PANAS-P had not reached significance (p = 0.337).

The analyses of heart rate variability and low blood pressure variability found no variance homogeneity.The analysis of high blood pressure variability revealed a significant effect of TIME, F(1, 52) = 4.926, p = 0.031, η² = 0.087. In details, the variability of high blood pressure in the pre-test (M_Pre = 4.07%, SD = 7.06%) was significantly higher than the variability of high blood pressure in the post-test (M_Pre = 0.60%, SD = 6.55%). However, we found no significant effect of GROUP and the interaction of GROUP and TIME on high blood pressure variability (see Fig. 3). All the comparation data and ANOVA results were listed in supplementary materials (see Tables 2 and 3). Additionally, no participants reported physical discomfort experienced after viewing virtual reality imagery.

INSERT FIGURE 3 ABOUT HERE

The one-way ANOVA analysis (n = 21) showed a significant main effect of TIME on LSAS scores (F(1, 28) = 18.224, p < 0.001, η² = 0.477) and SPIN scores (F(1, 28) = 26.519, p < 0.001, η² = 0.570). The post-hoc analyses revealed that the post-intervention scores (M_SPINpost = 29.25, SD = 9.26; M_LSASpost = 58.48, SD = 15.28) were significantly decreased than pre-intervention scores (M_SPINpre = 40.00, SD = 11.41; M_LSASpre = 69.90, SD = 20.53), as well as the follow-up scores (M_{SPINfollow−up} = 26.62, SD = 8.90; M_{LSASfollow−up} = 50.19, SD = 17.66) were significantly decreased than pre-intervention scores (all ps < 0.05).

We also found a significant main effect of TIME on CORE-OM scores (F(1, 23) = 8.459, p < 0.01, η² = 0.297), PANAS-P scores, and PANAS-N scores. The post-hoc analyses revealed that the post-intervention scores (M_{CORE−OMpost} = 40.86, SD = 10.97; M_{PANAS−Ppost} = 29.05, SD = 5.05; M_{PANAS−Npost} = 21.86, SD = 5.43) are significantly improved than pre-intervention scores (M_CORE−OMpre = 47.62, SD = 14.88; M_PANAS−Ppre = 26.90, SD = 4.55; M_PANAS−Npre = 28.05, SD = 6.88), as well as the follow-up scores (M_{CORE−OMfollow−up} = 36.95, SD = 10.39; M_{PANAS−Pfollow−up} = 29.24, SD = 6.00; M_{PANAS−Nfollow−up} = 20.38, SD = 5.47) are significantly improved than pre-intervention scores (all ps < 0.05).

The main effects of TIME on high and low blood pressure variability reached significance (high blood pressure variability: F(1, 20) = 8.451, p < 0.01, η² = 0.297; low blood pressure variability: F(1, 20) = 18.742, p < 0.001, η² = 0.484). The post-hoc analyses revealed that the post-intervention scores (M_HPVpost = 1.05%, SD = 6.59%; M_LPVpost = -3.63%, SD = 7.31%) are significantly improved than pre-intervention scores (M_HPVpre = 5.20%, SD = 6.90%; M_LPVpre = 2.31%, SD = 6.49%) (see Fig. 4). All the comparation data and ANOVA results were listed in supplementary materials (see Tables 4 and 5).

INSERT FIGURE 4 ABOUT HERE

The primary outcome measure in this study is the level of social anxiety. After 14 days of self-help intervention, the intervention group showed a significant decrease in total scores on the SPIN scale, particularly in the fear and avoidance dimensions, while the waitlist group did not show the same significant results. Similarly, the intervention group also exhibited significant reductions in total scores and across the fear, anxiety, and avoidance dimensions on the LSAS scale, in contrast to the waitlist group which did not demonstrate significant changes. Based on these findings, the study concludes that the self-help intervention with convenient VR appliances effectively reduces self-report social anxiety. This finding is consistent with past studies showing the efficacy of VRET for social anxiety (Kampmann et al., 2016; Kim et al., 2022). Specifically, our study used a mobile-based self-training VR program without any therapist assistances, which provided self-training VRET outcome evidence in a Chinese context.

Comparing with previous studies, our study showed superior intervention outcomes. For example, Kim et al. (2017) used self-training mobile-based VR program and found that patients showed greater degree of decrease at the marginal significance level than controls (p = 0.053). In their study, SA participants were screened using the Mini International Neuropsychiatric Interview module (MINI) to determine the presence of SAD; however, we used self-report scales (e.g., LSAS and SPIN) as including criterion, which might contribute to the self-report improvements on symptoms. Kampmann et al. (2016) investigated the efficacy of therapist-lead VRET (ten 90 min sessions scheduled twice a week) and found a significant decrease from pre- to post-assessment for VRET (p = .014; d = 0.55) compared to the waiting-list control group. In our study, the self-help training group had a significant decrease on the LSAS scores (p < 0.001, η² = 0.348), which had better cost-effectiveness.

Additionally, positive and negative emotions were served as important indicators in this self-help training program. After 14 days of self-help intervention, the intervention group showed a significant decrease in scores on the negative emotion dimension of the PANAS scale, while the waitlist group did not show the same significant results. However, there were no significant differences in changes in scores on the positive emotion dimension between the two groups. In fact, the training content appears to reduce inner negative feelings, which generally aligns with existing research findings. Regarding the positive correlation between negative emotions and social anxiety, previous studies have indeed shown that social anxiety disorder is highly comorbid with negative emotions such as depression. The correlation between the two can be measured using the PANAS scale (e.g., Hughes & Kendall, 2009). Moreover, exposure therapy has been shown to effectively intervene in the negative evaluation and fear of unknown social situations (Anderson et al., 2013; Campbell-Sills et al., 2015).

Within the intervention group, after the 14-day self-help intervention, significant reductions were observed in total scores and most dimension scores on the SPIN and LSAS scales, and these effects were sustained during the one-month follow-up. In the waitlist group, there were no significant differences in total scores or most dimension scores on the SPIN and LSAS scales between the pre-waiting and post-waiting periods, but significant reductions were observed during the intervention period, and these effects were also sustained during the one-month follow-up. The within-subject design of social anxiety improvement provides further long-term evidence for the randomized controlled trial results.

Additionally, the most common physiological indicators are cardiovascular reactivity, primarily including heart rate variability and blood pressure variability. Some studies found that individuals with high social anxiety exhibit elevated cardiovascular reactivity when facing social tasks compared to those with low social anxiety (e.g., Liu et al., 2023; Lü et al., 2022). In our study, we found the variability of high blood pressure in the pre-test was significantly higher than the variability of high blood pressure in the post-test. However, we did not find interaction effect of GROUP and TIME. However, some previous studies found a decrease in heart rate variability reactivity after psychological interventions (e.g., Kim et al., 2017; Sülter et al., 2022). These decreases might suggest that participants both in the intervention group and waiting control group were less stressed by the anticipation of repeated social interaction with the experimenters.

The study findings should be considered in light of certain limitations. First of all, live-action shooting is limited by the number of actors, and only up to six characters can appear in one scene. However, many roles can be easily realized through 3D animated characters. For example, an intervention study with public speaking anxiety set the number of variables as 10 people in small venues, 18 people in medium venues, and 46 people in large venues (Takac et al., 2019). In fact, this study did not touch on a real large public speaking scenario, and it must be acknowledged that there are difficulties in using real people to create videos for this social anxiety subtype. Secondly, the various positive or negative reactions of the video characters in this study are preset and cannot be adjusted independently. The flexible adjustment of stimulation levels is indeed inferior to VR videos presented with 3D animation (Lindner et al., 2021). Thirdly, regarding the tracking time, there is a staggeringly longest period of 6 years abroad (Wong et al., 2023). Due to time pressure, our study only formulated a one-month follow-up plan, which is really difficult to achieve. Fourthly, it should be noticed that besides self-rating scales and physiological indicators, there could be some theoretical methods—for example, behavioral scores—to gauge the improvement effect of social anxiety. Fifth, although the current meta-analysis did not focus on comparing the effect of VRET with a placebo, further comparing it with other psychological treatments might inflate the effect size of the advantage by merely comparing it with the waitlist controls. In the future, placebo controls would be worth adding to confirm the placebo effect of self-intervention. The proportion of female participants in this study was more than 80%. Although the proportion in 13 of the 15 existing VRET RCTs for social anxiety was between 60% and 79% (Wong et al., 2023) and meta-analysis studies never discussed the impact on interventional results (Carl et al., 2019; Chesham et al., 2018; Fodor et al., 2018; Hinojo-Lucena et al., 2020; Horigome et al., 2020; Powers & Emmelkamp, 2008; Wong et al., 2023), extreme proportion should still be avoided in the research designs. It involves more people, provides third-party evaluation, and the subject's behavioral performance in a situation truer to life; these are irreplaceable advantages.

To the best of one's knowledge, this was the first RCT which proved that self-help training VERT is an effective intervention for social anxiety in Chinese college students. It generalized to decreased adverse effects and improved mental health through exposure to convenient VR appliances. Reductions of social anxiety were maintained across both measures at one-month follow-up. Further clinical practice or research using self-help training VRET might need to consider various social scenes for specific populations of social anxiety.

Author Contribution

Chinghsiang Kan and Ya Zhang wrote the main manuscript text; Yifan Wang and Ruoyun Hu prepared figures and tables. Ke Chen contributed to the intervention process and grammar checking. All authors reviewed the manuscript.

Data Availability

Data is provided within the manuscript or supplementary information files

American Psychiatric Association. (2022). Diagnostic and statistical manual of mental disorders (5th ed.). https://doi.org/10.1176/appi.books.9780890425787
Anderson, P. L., Price, M., Edwards, S. M., Obasaju, M. A., Schmertz, S. K., Zimand, E., & Calamaras, M. R. (2013). Virtual reality exposure therapy for social anxiety disorder: A randomized controlled trial. Journal of Consulting and Clinical Psychology, 81(5), 751-760. https://doi.org/10.1037/a0033559
Beck, A. T., Steer, R. A., & Brown, G. K. (1996). BDI–II, Beck Depression Inventory: Manual (2nd ed.). Psychological Corp.
Bradley, H. A., Rucklidge, J. J., & Mulder, R. T. (2017). A systematic review of trial registration and selective outcome reporting in psychotherapy randomized controlled trials. Acta Psychiatrica Scandinavica, 135(1), 65-77. https://doi.org/10.1111/acps.12647.
Campbell-Sills, L., Espejo, E., Ayers, C. R., Roy-Byrne, P., & Stein, M. B. (2015). Latent dimensions of social anxiety disorder: A re-evaluation of the Social Phobia Inventory (SPIN). Journal of Anxiety Disorders, 36, 84-91. https://doi.org/10.1016/j.janxdis.2015.09.007
Carl, E., Stein, A. T., Levihn-Coon, A., Pogue, J. R., Rothbaum, B., Emmelkamp, P., Asmundson, G. J. G., Carlbring, P., & Powers, M. B. (2019). Virtual reality exposure therapy for anxiety and related disorders: A meta-analysis of randomized controlled trials. Journal of Anxiety Disorder, 61, 27-36. https://doi.org/10.1016/j.janxdis.2018.08.003.
Chen, H., Rodriguez, M. A., Qian, M., Kishimoto, T., Lin, M., & Berger, T. (2020). Predictors of treatment outcomes and adherence in Internet-based cognitive behavioral therapy for social anxiety in China. Behavioural and Cognitive Psychotherapy, 48(3), 291-303. https://doi.org/10.1017/S1352465819000730.
Chesham, R. K., Malouff, J. M., & Schutte, N. S. (2018). Meta-Analysis of the Efficacy of Virtual Reality Exposure Therapy for Social Anxiety. Behaviour Change, 35(3), 152-166. https://doi.org/10.1017/bec.2018.15.
Connor, K. M., Davidson, J. R. T., Churchill, L. E., Sherwood, A., Weisler, R. H., & Foa, E. (2000). Psychometric properties of the Social Phobia Inventory (SPIN): New self-rating scale. British Journal of Psychiatry, 176(4), 379-386. https://doi.org/10.1192/bjp.176.4.379
Ding, X.F., & Li, D. (2018). Meta-Analysis of Randomized Controlled Trials of Virtual Reality Exposure Therapy for Anxiety Disorders. Chinese Journal of Mental Health, 32(03), 191-199. (in Chinese). https://doi:10.3969/j.issn.1000-6729.2018.03.003
Evans, J. M.-C. F. M. M. B. K. A. J. C. G. M. C. (2000). CORE: Clinical Outcomes in Routine Evaluation. Journal of Mental Health, 9(3), 247-255. https://doi.org/10.1080/jmh.9.3.247.255.
Garcia-Palacios, A., Botella, C., Hoffman, H., & Fabregat, S. (2007). Comparing acceptance and refusal rates of virtual reality exposure vs. in vivo exposure by patients with specific phobias. Cyberpsychological Behaviour, 10(5), 722-724. https://doi.org/10.1089/cpb.2007.9962.
He, Y.L., & Zhang, M.Y. (2004). The Reliability and Validity Study of the Liebowitz Social Anxiety Scale. Diagnosis: Theory and Practice, 03(02), 29-33. (in Chinese). http://doi.org/10.3969/j.issn.1671-2870.2004.02.009.
Hinojo-Lucena, F.-J., Aznar-Díaz, I., Cáceres-Reche, M.-P., Trujillo-Torres, J.-M., & Romero-Rodríguez, J.-M. (2020). Virtual Reality Treatment for Public Speaking Anxiety in Students. Advancements and Results in Personalized Medicine. Journal of personalized medicine, 10(1), 14. https://doi.org/10.3390/jpm10010014
Horigome, T., Kurokawa, S., Sawada, K., Kudo, S., Shiga, K., Mimura, M., & Kishimoto, T. (2020). Virtual reality exposure therapy for social anxiety disorder: a systematic review and meta-analysis. Psychological Medicine, 50(15), 2487-2497. https://doi.org/10.1017/S0033291720003785
Huang, L., Yang, T.Z., & Ji, Z.M. (2003). Research on the Applicability of the Positive and Negative Affect Schedule in the Chinese Population. Chinese Journal of Mental Health, 17(01), 54-56 (in Chinese). https://doi.org/10.3321/j.issn:1000-6729.2003.01.018
Hughes, A. A., & Kendall, P. C. (2009). Psychometric Properties of the Positive and Negative Affect Scale for Children (PANAS-C) in Children with Anxiety Disorders. Child Psychiatry & Human Development, 40(3), 343-352. https://doi.org/10.1007/s10578-009-0130-4
Kampmann, I. L., Emmelkamp, P. M., Hartanto, D., Brinkman, W. P., Zijlstra, B. J., & Morina, N. (2016). Exposure to virtual social interactions in the treatment of social anxiety disorder: A randomized controlled trial. Behaviour research and therapy, 77, 147-156. https://doi.org/10.1016/j.brat.2015.12.016
Keller, M. B. (2003). The lifelong course of social anxiety disorder: a clinical perspective. Acta Psychiatrica Scandinavica, 108(417), 85-94. https://doi.org/10.1034/j.1600-0447.108.s417.6.x
Kennedy, R. S., Lane, N. E., Berbaum, K. S., & Lilienthal, M. G. (1993). Simulator Sickness Questionnaire: An Enhanced Method for Quantifying Simulator Sickness. International Journal of Aviation Psychology, 3(3), 203-220. https://doi.org/10.1207/s15327108ijap0303_3
Kim, H. E., Hong, Y.-J., Kim, M.-K., Jung, Y. H., Kyeong, S., & Kim, J.J. (2017). Effectiveness of self-training using the mobile-based virtual reality program in patients with social anxiety disorder. Computers in Human Behavior, 73, 614-619. https://doi.org/10.1016/j.chb.2017.04.017
Kim, M. K., Eom, H., Kwon, J. H., Kyeong, S., & Kim, J. J. (2022). Neural effects of a short-term virtual reality self-training program to reduce social anxiety. Psychol Med, 52(7), 1296-1305. https://doi.org/10.1017/S0033291720003098
Kothgassner, O. D., & Felnhofer, A. (2021). Lack of research on efficacy of virtual reality exposure therapy (VRET) for anxiety disorders in children and adolescents : A systematic review. Neuropsychiatr, 35(2), 68-75. https://doi.org/10.1007/s40211-020-00349-7
Liebowitz, M. R. (1987). Social phobia. Modern problems of pharmacopsychiatry, 22, 141-173. https://doi.org/10.1159/000414022
Liebowitz, M. R., Heimberg, R. G., Schneier, F. R., Hope, D. A., Davies, S., Holt, C. S., Goetz, D., Juster, H. R., Lin, S. H., Bruch, M. A., Marshall, R. D., & Klein, D. F. (1999). Cognitive-behavioral group therapy versus phenelzine in social phobia: Long-term outcome. Depression and Anxiety, 10(3), 89-98. https://doi.org/10.1002/(SICI)1520-6394(1999)10:3<89::AID-DA1>3.0.CO;2-5
Lindner, P., Dagoo, J., Hamilton, W., Miloff, A., Andersson, G., Schill, A., & Carlbring, P. (2021). Virtual Reality exposure therapy for public speaking anxiety in routine care: a single-subject effectiveness trial. Cogn Behav Ther, 50(1), 67-87. https://doi.org/10.1080/16506073.2020.1795240
Lindner, P., Miloff, A., Hamilton, W., Reuterskiold, L., Andersson, G., Powers, M. B., & Carlbring, P. (2017). Creating state of the art, next-generation Virtual Reality exposure therapies for anxiety disorders using consumer hardware platforms: design considerations and future directions. Cogn Behav Ther, 46(5), 404-420. https://doi.org/10.1080/16506073.2017.1280843
Liu, X., Lin, W., Huang, J., Cao, Z., Wu, M., Chen, Z., Zhu, W., Tan, Z., Yu, P., Ma, J., Chen, Y., Zhang, Y., & Wang, J. (2023). Depressive symptoms, anxiety and social stress are associated with diminished cardiovascular reactivity in a psychological treatment-naive population. Journal of Affective Disorders, 330, 346-354. https://doi.org/10.1016/j.jad.2023.02.150
Lü, W., Zhu, W., & Wang, S. (2022). Associations between general trait anxiety, trait social anxiety and cardiovascular responses to social and non-social stressors. Journal of Psychosomatic Research, 160, 110979. https://doi.org/10.1016/j.jpsychores.2022.110979
Meyerbröker, K., & Emmelkamp, P. M. G. (2010). Virtual reality exposure therapy in anxiety disorders: a systematic review of process-and-outcome studies. Depression and Anxiety, 27(10), 933-944. https://doi.org/10.1002/da.20734
Poeschl, S., & Doering, N. (2014). Effects of simulation fidelity on user experience in virtual fear of public speaking training:An experimental study. Annual Review of CyberTherapy and Telemedicine, 12, 66-70.
Prochaska, J., & Norcross, J. (2018). Systems of Psychotherapy: A Transtheoretical Analysis, 9th ed. Wadsworth Pub Co.
Ruscio, A. M., Brown, T. A., Chiu, W. T., Sareen, J., Stein, M. B., & Kessler, R. C. (2008). Social fears and social phobia in the USA: results from the National Comorbidity Survey Replication. Psychol Med, 38(1), 15-28. https://doi.org/10.1017/S0033291707001699
Safir, M. P., Wallach, H. S., & Bar-Zvi, M. (2012). Virtual reality cognitive-behavior therapy for public speaking anxiety: one-year follow-up. Behav Modif, 36(2), 235-246. https://doi.org/10.1177/0145445511429999
Schulz, K. F., Altman, D. G., & Moher, D. (2010). CONSORT 2010 statement: Updated guidelines for reporting parallel group randomized trials. Annals of Internal Medicine, 152(11). https://doi.org/10.7326/0003-4819-152-11-201006010-00232.
Sevinc, V., & Berkman, M. I. (2020). Psychometric evaluation of Simulator Sickness Questionnaire and its variants as a measure of cybersickness in consumer virtual environments. Applied Ergonomics, 82, 102958. https://doi.org/10.1016/j.apergo.2019.102958
Seyama, J. i., & Nagayama, R. S. (2007). The Uncanny Valley: Effect of Realism on the Impression of Artificial Human Faces. Presence : teleoperators and virtual environment, 16(4), 337-351. https://doi.org/10.1162/pres.16.4.337
Sülter, R. E., Ketelaar, P. E., & Lange, W.G. (2022). SpeakApp-Kids! Virtual reality training to reduce fear of public speaking in children – A proof of concept. Computers & Education, 178, 104384. https://doi.org/10.1016/j.compedu.2021.104384.
Takac, M., Collett, J., Blom, K. J., Conduit, R., Rehm, I., & De Foe, A. (2019). Public speaking anxiety decreases within repeated virtual reality training sessions. PLoS One, 14(5), e0216288. https://doi.org/10.1371/journal.pone.0216288
Xiao, R., Wu, W.L., & Zhang, W. (2007). The Application of the Liebowitz Social Anxiety Scale among High School Students in Chengdu City. West China Medical Journal, 22(04), 708-710. (in Chinese). http://doi.org/10.3969/j.issn.1002-0179.2007.04.007
Wallach, H. S., Safir, M. P., & Bar-Zvi, M. (2009). Virtual reality cognitive behavior therapy for public speaking anxiety: A randomized clinical trial. Behavior Modification, 33(3), 314-338. https://doi.org/10.1177/0145445509331926
Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative affect: The PANAS scales. Journal of Personality and Social Psychology, 54(6), 1063-1070. https://doi.org/10.1037/0022-3514.54.6.1063
Wong, K. P., Lai, C. Y. Y., & Qin, J. (2023). Systematic review and meta-analysis of randomised controlled trials for evaluating the effectiveness of virtual reality therapy for social anxiety disorder. Journal of Affective Disorders, 333, 353-364. https://doi.org/10.1016/j.jad.2023.04.043
World Health Organization. (2019). International statistical classification of diseases and related health problems (11th ed.). https://icd.who.int/
Ye, Z.J., Wang, Z.J., Liang, M.Z., Gao, Y.Y., Yu, Y.L. et al.(2018). Advancements in the Diagnosis and Treatment of Social Anxiety Disorder. Medicine and Philosophy,39(04), 56-59. (in Chinese). http://doi.org/10.12014/j.issn.1002-0772.2018.04b.16.
Zaboski, B. A., Joyce-Beaulieu, D., Kranzler, J. H., McNamara, J. P., Gayle, C., & MacInnes, J. (2019). Group exposure and response prevention for college students with social anxiety: A randomized clinical trial. J Clin Psychol, 75(9), 1489-1507. https://doi.org/10.1002/jclp.22792
Zhang, Y., Hu, J., Jin, L.W., Wu, M.Y., Wang, C.Y. et al. (2020). Psychometric properties of the Chinese version of the clincial outcomes in routine evaluation-outcome measure (CORE-OM). British Journal of Guidance and counseling, 48, 289-299. https://doi.org/10.1080/03069885.2019.1682120.
Zhang, X.X., Ke, X.X., Zhang, L.R., Jiang, J.H., & Wang, Y.M. (2002). The Validity and Reliability Assessment of the Simulator Sickness Questionnaire (Chinese Version) in Evaluating College Students in Simulated Environments. Chinese Journal of Mental Health, 36(01), 68-72. http://doi.org/10.3969/j.issn.1000-6729.2022.01.011

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Convenient virtual reality exposure self-training for social anxiety: a randomized controlled study

Status:

Version 1

Abstract

Figures

Introduction

Methods

INSERT FIGURE 1 ABOUT HERE

Measures

VR apparatus & Stimuli

INSERT FIGURE 2 ABOUT HERE

Procedure

Analysis

Result

INSERT Table ABOUT HERE

INSERT FIGURE 3 ABOUT HERE

INSERT FIGURE 4 ABOUT HERE

Discussion

Declarations

Author Contribution

Data Availability

References

Additional Declarations

Supplementary Files

Status:

Version 1