Virtual reality (VR) is being used in various fields of medicine along with the development of head-mounted display (HMD) devices (1–3). In psychiatry, VR has been mainly used to treat patients with specific phobia or posttraumatic stress disorder (PTSD) (4–8). Recently, VR has been widely used to treat various other mental disorders (9–11). However, since VR can cause cyber sickness, caution is needed to use VR for therapeutic intervention (12). Cyber sickness is a collection of symptoms that occur during exposure to VR immersion, including eye strain, headache, vertigo, ataxia, sweat, disorientation, and dizziness (13–15). To improve treatment efficacy for patients with anxiety, it is necessary to investigate the increase in anxiety caused by cyber sickness when watching VR videos.

Skin conductance (SC) is widely used as a bio-signal reflecting anxiety because it is positively correlated with anxiety (16–18). SC, also referred to as electrodermal activity (EDA), is the measurement of the electrical conductivity of the skin (19). It has been found that depression is associated with a decrease in EDA (20), while fear and anxiety are associated with an increase in EDA (21). The EDA level (tensile level of electrical conductivity) is measured by placing two electrodes in a dense position of the eccrine sweat gland. It is typically measured for fingers or hands (22). However, there are variations depending on the user's skin condition. A high value is measured in the case of a sweaty constitution, while a low value is measured in the case of severe dry skin.

Based on the fact that shaking of the video causes cyber sickness (13–15), the purpose of this study was to determine whether the order of exposure to the VR image with increasing or decreasing the degree of dizziness was associated with participant's anxiety and whether such anxiety could be measured in real time based on skin conductance. The association between anxiety and skin conductance was analyzed by varying the exposure order of VR images. In treatment with virtual reality, it will be important to know whether the most dizzy substance should be introduced at the end of treatment or at the beginning of treatment to reduce cyber sickness. This study can help solve the problem. We hypothesized that exposure to dizzy images would increase anxiety and skin conductivity compared to baseline.

Demographic characteristics were not significantly different between the two groups except for pulse rate (p < 0.027). As shown in Table 1, mean age was 39.98 years for the Order 1 group and 36.54 years (p < 0.205) for the Order 2 group. In the Order 1 group, 45% were male patients and 55% were female patients. In the Order 2 group, 51.22% were male patients and 48.78% were female patients (p < 0.575). The percentage of those without motion sickness was 72.5% in the Order 1 group and 85.37% in the Order 2 group (p < 0.155).

As shown in Table 2, SC was increased significantly in both groups after VR video exposure than that at baseline (p < 0.0001). Mean (±SD) difference of SC between the first VR and the baseline was 1.54 (±1.17) for the Order 1 group and 0.85 (±0.86) for the Order 2 group. Changes in SC in the two groups between the first VR and the baseline were statistically significant (p < 0.005). Mean (±SD) difference of SC between the second VR and the baseline was 1.61 (±1.07) for the Order 1 group and 0.92 (±0.9) for the Order 2 group. Changes in SC in the two groups between the second VR and the baseline were statistically significant (p < 0.003). The difference (mean ± SD) of SC between the second VR and the first VR in the Order 1 group was 0.07 ± 0.47, which was statistically significant (p < 0.036). STAI-X-1 was decreased significantly after the first VR exposure (p < 0.001) and the second VR exposure (p < 0.035) as compared to baseline in the Order 2 group. Mean (±SD) change of STAI-X-1 in the Order 2 group was −4.24 (±6.66) between the first VR and baseline and −3.12 (±9.16) between the second VR and baseline. Changes in STAI-X-1 between the first VR and the baseline were statistically significant between the two groups (p < 0.031).

As shown in Table 3, skin conductance difference after baseline correction was not significantly different between the two groups (p >0.05). Mean (± SD) difference of SC for Day 1 PM between the second VR and baseline was 3.23 (±3.6) for the Order 1 group and 2.2 (±2.41) for the Order 2 group.

All measurements of skin conductance by time point and the line connecting the average time point are displayed in a box-whisker plot (Figure 2). Skin conductance was found to be steeper in the group that watched more dizzy VR than that in the group that watched less dizzy VR (Order 2) in the first virtual reality stimulus. Measurement of skin conductance was flat regardless of the intensity of dizziness in the second VR stimulus.

We predicted that anxiety would gradually increase after exposure to a dizzy video compared to that at baseline. However, the results were somewhat different from our prediction. It might be due to the following two reasons. First, baseline anxiety was highly measured in both groups due to tension caused by the first experience of experimental environment. The second reason was that the dizzy video produced was based on natural scenery images. Thus, anxiety was measured to be lower than expected. In addition, the VR image was used based on what could be used to treat patients with high anxiety and minimize the burden on participants. However, it was meaningful that anxiety was slightly increased in the Order 1 group after the last video exposure, while anxiety was decreased statistically significantly in the Order 2 group (p < 0.035). This was in line with an increase in higher skin conductivity change rate in the Order 1 group after exposure to the last video (mean ± SD of SC difference between second VR and baseline: 1.61 ± 1.07). It could be interpreted that the increase in anxiety was greater when watching less dizzy videos after watching strong dizziness videos than in the opposite case.

The amount of change in skin conductance between the two groups did not show a significant difference after correcting the baseline. This might be due to the following reasons. First, pulse rate of the Order 1 group was significantly (p < 0.027) higher than that of the Order 2 group. This might have led to highly measured anxiety and skin conductance at baseline. Second, the number of participants was small. Skin conductance reacts sensitively to a sweaty constitution. If the skin is severely dry, there is a deviation such as a low measurement value. If the sample size is small, the deviation might be greater (21, 34). As shown in Table 3, from Day 1 AM to Day 2 PM, the average change of SC between first VR and baseline decreased from 3.22 to 1.1 in the Order 1 group and from 2.13 to 1.22 in the Order 2 group, showing a tendency to decrease. However, when the date and time were corrected, the difference between the two groups was insignificant (p = 0.2049). This indicates that when exposed to a more dizzy VR for the first time, the measurement of skin conductivity increases. When one gets used to VR later, there is no significant difference in the measurement regardless of which VR is watched first.

In conclusion, our study showed that anxiety and skin conductivity increased more when watching less dizzy video after watching a more dizzy video than in opposite situations. This might help us prepare image composition when VR is used for treatment intervention to reduce anxiety in future. It is possible to reduce skin conductance and anxiety by viewing less dizzying VR videos first and then viewing more dizzying videos later.

VR with HMD has the advantage of maximizing immersion. However, using 360° images has the disadvantage of causing cyber sickness depending on the degree of stabilization of the camera (35). In our study, it was confirmed that skin conductance was significantly increased after watching the last VR video. Since cyber sickness can cause headache, vomiting, and dizziness, attention should be paid to treatment intervention (36, 37). Therefore, when virtual reality is used for therapeutic intervention, a careful approach is needed. Previous studies have shown that breathing training is effective in reducing dizziness (38). In addition, considering the order of exposure to dizziness, it will be possible to make more effective settings when using virtual reality for therapeutic intervention. A more complete setting such as preparing images and environments to minimize cyber sickness is needed when VR is used for relaxation training of patients with depression and anxiety in future. However, caution is needed when generalizing the results of this study as this study was conducted only with Korean people. More studies enrolling subjects from various cultures are needed to verify our results.

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author.

The studies involving humans were approved by Institutional Review Board of Samsung Medical Center. The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.

HK, KK, and HJ contributed to conception and design of the study. M-JK performed the statistical analysis. DK wrote the first draft of the manuscript. All authors contributed to manuscript revision, read, and approved the submitted version.