A total of six healthy volunteers, four men and two women, aged 23–45 years, all Russian, with no or minimal previous experience in space simulations with isolation, were selected by the IBMP Medical Board.

To preserve the confidentiality of the subjects, each of them is designated by a code (A, B, C, D, E, and F).

The studies involving human participants were reviewed and approved by the Bioethical Commission of the Institute of Biomedical Problems of the Russian Academy of Sciences and fully complied with the principles of the 1964 Declaration of Helsinki.

Each study participant voluntarily signed informed consent after having the potential risks, benefits, and nature of the upcoming study explained to them.

The ESKIS (experiment with short-term isolation, which when translated into Russian gives “ЭСКИЗ–Эксперимент С Короткой Изоляцией,” lit. “sketch”) experiment with 14-day isolation and crowding of the mixed-gender crew was carried out in a chamber of 50 cubical meters (18 m² of living space) at the SSC RF-IMBP RAS (Nosenkova, 2021). This habitable volume is comparable to the living space of orbital stations (Salyut-type stations have a volume of 90.0 m³, the Destiny ISS module has a volume of 117 m³, and the International Habitat Module of the upcoming Lunar Gateway station will have a volume of 125 m³). However, spaceships and modules of orbital stations comparable to the habitable volume of the module in the ESKIS experiment should be considered psychologically more spacious, since astronauts can use all available surfaces due to weightlessness. The set of furniture in the ESKIS experiment was strictly limited and included two folding tables for work and eating, folding chairs, and a stationary table with equipment for studies and communication with Mission Control (Figure 1). The number of berths in the module was less than the number of crew members, so two additional sleeping places were arranged in the aisle.

The daily routine was specified in accordance with the experiment program, including execution of the operator’s performance tasks, three meals, 8 h of sleep, a medical supervision program, and experimental studies.

The main psychological factors affecting the subjects during the experiment are typical for chamber confinement studies (as well as isolation caused by epidemiological reasons): sensory deprivation, monotony, crowding, lack of privacy, and the limitation of social contacts (Feichtinger et al., 2012).

For statistical analysis, the Wilcoxon t-test, as well as the Spearman and Pearson correlation, was used, with consideration of the sample sizes, types and distribution of the variables, and analysis goals. The analysis was carried out using SPSS software.

It should be noted that the small sample size (considering the number of subjects) poses a significant limitation to the present study, although the experimental design allowed frequent data collection for each subject.

Analysis of the negative mood states (tension or anxiety, anger or hostility, fatigue or inertia, depression or dejection, and confusion or bewilderment) by applying POMS before, during, and after the isolation period revealed the increase in negative feelings before the chamber study, on the first day of it, and after the isolation period for most of the crew. For two subjects, D and F, some increase in negative emotions was also detected in the middle of the mission. However, in general, no tendencies for the increase in tension, anxiety, and depression were found throughout isolation (Figure 2). More or less, the same results were obtained using the STAI test; a pronounced increase was detected only before and during the first day of isolation. These changes could be attributed to the preparatory stage and acute period of adaptation to the experimental conditions.

The dynamics of the duration of morning and evening DPCs also demonstrated only two major spikes on 2–3 days of isolation and after the simulation of lunar extra-vehicular activity (EVA) when crew members moved in pairs to a bigger chamber and performed for 45 min a number of complex operator’s tasks (Figure 3). In general, the duration of the evening video reports decreased (p = 0.043).

The analysis of the expression of basic emotions during DPCs demonstrated that for four subjects, the higher was the duration of the DPC, the lower was the expression of happiness (p = 0.023), surprise (p < 0.001), valence (p < 0.001), and the level of excitement (p < 0.001). At the same time, sadness (p = 0.019) and anger (p = 0.008) increased in this group during the isolation period. For two subjects, no dynamics in emotional expression was found. For subject D, the tendency for growing neutral emotions in the facial expression was found (p = 0.015). These data correspond to the previous results from longer isolation studies MARS-500 and SIRIUS-19 (Supolkina et al., 2021).

Some correlations between acoustic parameters, such as the number of pauses, speech loudness, and shimmer effect (variation of the voice amplitude), during DPCs and the STAI anxiety level were revealed. Аnxiety levels were negatively correlated with speech loudness and positively correlated with the number of pauses and shimmer effect (Table 1). Also, we found that the increase in the anxiety level was positively correlated with the number of errors and time for mathematical calculation in cognitive tests.

The study of motor activity under confinement and crowding showed an expected decrease of this parameter. Motor activity in the small 50-m³ chamber was 30% lower than that during the preparatory stage and 29% lower than that during the recovery period.

The use of actigraphy made it possible to assess the duration and quality of sleep (including 3 days of the baseline period and 14 days of isolation). For the whole crew, except B, the decrease in sleep duration was observed (Figure 4). All subjects, except B, had a tendency to decrease the amount of time spent in bed which was defined by analyzing actigraphy data using the Cole–Kripke algorithm (Figure 5). The number of night awakenings for subjects A and D increased in the course of isolation (Figure 6). In general, sleep duration and quality were negatively affected by experimental conditions.

Analysis of the value–orientation unity in the group demonstrated that for the positive traits (values), the subjects, in general, showed similar dynamics during isolation, except for two of them (Figure 7A). At the same time, for the values which the group regarded as unacceptable under isolation, subject B demonstrated negative dynamics, while the other subjects became psychologically closer (Figure 7B). Analysis of the common goals (subject–goal unity) produced similar results (Figure 8). In the group, five members became close in the goals to reach during isolation and one of them demonstrated increasing inconsistency.

Data on the perceived psychological similarity defined by the PSPA test is shown in Table 2. The bigger the distance in the semantic space between a subject’s “Real Ego” and the other crew members, the bigger will be the perceived difference, and vice versa. During isolation, we detected statistically non-significant (according to the Wilcoxon t-criterion) tendency for the decrease in distances, which could be interpreted as the increase in the perceived similarity throughout their stay in the chambers. It is important to note that during the second testing in isolation (MD13), as well as after isolation, we detected statistically significant differences between “Ideal Ego” and perceived images of other subjects (р = 0.028).

In general, the sociometric test, a study of group values and perceived images of the crew members, shows the increase in crew cohesion during isolation (Figure 9). Only one member of the group seemed to be not compatible with the rest.

Results obtained with the KTS are shown in Table 3. It is evident that almost all subjects demonstrated evident dominance of judging. Three of them (A, D, and E) also gave preference to sensing and thinking, so they can be attributed to the sensory-planning type. Subject B, who, as we found out, did not share the crew’s common values, demonstrated dominance of intuition. Furthermore, we correlated the subjects’ types of processing information with their preferences to types of psychological support.

During the preparatory stage of the experiment, all subjects gained information about the opportunity to install their own video and audio content, including photos and books, on their personal tablet computers. This content could be used during their leisure time.

To provide medical safety criteria, individual intolerance to VR utilization as a means of psychological support was defined by the interview and special neurological tests in sitting and standing positions. Two subjects complained about symptoms of motion sickness (vertigo and nausea) in their common life. Expanded neurological examination defined that they had nystagmus after a VR session. No other neurological symptoms like tremor or instability during Romberg’s test were found. During the isolation period, these subjects did not complain about any vestibular problems during VR sessions.

According to the “psychological support” questionnaire, the average values of “PS from communication” were positively correlated both with the degree of association of the subjects with the crew (perception of themselves as part of the group) and with the level of support from the crew (ρ = 0.888, p < 0.05 and ρ = 0.858, p < 0.05, respectively). The “extraversion” parameter from KTS had a positive correlation with the indicator of psychological support from communication (i.e., social support) with the duty crew (ρ = 0.896; p < 0.05). At the same time, persons with a tendency for introversion preferred to use psychological support provided via virtual reality technology (ρ = −0.896, p < 0.05). This allows us to assume a direct relationship between the amount (volume) of communication in the crew and with the duty crews, a sense of unity (association) with the crew, and the favorable and supportive influence of this communication on the subjects with the dominance of extraversion.

In addition, it was shown that the indicator of the need for generalization and interpretation of information (KTS “thinking” scale) was negatively correlated with support from VR (ρ = −0.858, p < 0.05). Also, the need for the formal conservative pragmatic data and common sense (“sensation” scale, according to Keirsey) was positively correlated with the need for PS based on VR (ρ = 0.888, p < 0.05).

Not all the data from the video-recorded self-reports before and after VR sessions were available due to technical problems. The data describing basic emotions defined using Noldus FaceReader software via the subjects’ facial expressions showed that after VR sessions, the level of so-called neutral facial expression and «Anger» decreased, while the levels of «Happy», «Surprise», and «Arousal» increased. The detected parallel increase of «Disgust » contradicts the positive tendencies mentioned previously. Still, this result correlates with the previous data from the dry immersion study, where some subjects complained about the return of negative emotions immediately after leaving the virtual world and returning to the real stressful conditions (Rozanov et al., 2022). Also, similar to the dry immersion study, after VR sessions, we found a statistically significant decrease in cognitive mistakes (slips of the tongue and expletives) in self-reports (T = −2,101; p < 0.05). For subjects D, E, and F, after two sessions of VR-based PS, we also detected a statistically insignificant decrease in night awakenings and the increase in sleep duration.