Theoretical framework

Within the context of depression-related emotional dysfunction described earlier, Seligman’s well‑being framework provides a useful perspective for conceptualizing improvements in positive emotional and cognitive functioning. In Seligman’s theoretical framework (2011), well-being is characterized by the amalgamation of cognitive happiness (satisfaction), hedonic happiness (feeling), and eudaimonia (meaning). Well-being is proposed to depend on five key elements denoted as positive emotion, engagement, relationships, meaning, and achievement, succinctly represented by the acronym PERMA. The integration of these PERMA elements is posited to stimulate flourishing, signifying the optimal functioning of individuals, groups, communities, nations, and society as a whole. Augmenting the PERMA elements is proposed as a strategy to enhance overall well-being [47].

Building on this theoretical framework, this study examines the impact of VR-based positive psychotherapy on a triad of well-being indicators: life satisfaction (cognitive well-being), affect (hedonic well-being), and suicidal ideation (a critical marker of distress). These outcomes were selected not as isolated or competing endpoints, but as theoretically coherent components of psychological well-being that map onto distinct facets of the PERMA model [48]. Consistent with the aims of positive psychotherapy, the study focuses on multidimensional indicators of well‑being rather than a single symptom‑based clinical outcome. This approach provides an integrated and conceptually grounded assessment of well‑being, addressing concerns about multiplicity by focusing on complementary dimensions of a unified construct rather than unrelated clinical outcomes.

Quantitative phase

The quantitative phase of this research utilized a randomized controlled trial with a parallel-group design and a 1:1 allocation ratio. The procedures and documentation of all methodologies employed in the trial adhered to the CONSORT guidelines. A completed CONSORT checklist is provided as Supporting Information. Participant flow throughout the trial is presented in the CONSORT flow diagram (Fig 1).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. CONSORT flow diagram.

https://doi.org/10.1371/journal.pone.0354610.g001

Participants and settings

The participants in this study were adult patients referred to the psychiatric departments of two general hospitals situated in the southern region of Iran. All participants were diagnosed with unipolar MDD according to the criteria specified in the Fifth Edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-V) [50]. The diagnosis was additionally verified by a psychiatrist using the Structured Clinical Interview for DSM (SCID-5).

All participants continued routine pharmacological treatment during the study. All patients were prescribed sertraline, a selective serotonin reuptake inhibitor (SSRI) commonly used as a first-line treatment for MDD, within the usual therapeutic range (25–200 mg/day), as determined by the treating psychiatrist. Medication management was part of standard clinical care and was not manipulated by the research team. Participants were not required to initiate, discontinue, or modify their pharmacological treatment in order to participate in the study. Dosages remained stable during the brief intervention period, and the treating psychiatrist was not involved in delivering the psychological interventions or outcome assessments. No systematic differences in medication management were observed between groups; therefore, pharmacotherapy was not considered an experimental component of the study.

To meet the eligibility criteria for participation, participants had to be aged 18 or older, express willingness and preparedness to participate in the study, demonstrate literacy in reading and writing, have a psychiatrist-confirmed diagnosis of major depression, obtain a depression score exceeding 10 on the Patient Health Questionnaire (PHQ-9), report suicidal ideations as documented in the patient’s medical record, and be assessed through the Composite International Diagnostic Interview (CIDI) done by a senior psychiatric nursing student. Additionally, participants were required to have no psychotic or other mental disorders and no debilitating chronic or severe physical illnesses, as indicated in their medical records. This study employed the following exclusion criteria: individuals with psychotic or other mental health disorders other than MDD; debilitating chronic or severe physical illnesses, as indicated in their medical records, that could interfere with psychotherapy; those with a suicide risk requiring immediate hospitalization; individuals experiencing unexpected physical or mental problems during the study; and those unable to actively participate in the study process due to reasons such as slowness of thinking or speech. Additionally, participants expressing reluctance to continue contributing to the study were also excluded. All study procedures were conducted in accordance with the study protocol approved by the institutional ethics committee, which is provided as Supporting Information.

Randomization and blinding

Patients were randomly allocated to two groups: the intervention group, receiving positive psychotherapy via VR, and the control group, receiving traditional face‑to‑face positive psychotherapy. Randomization was performed using computer-generated allocation sequences. The numbers corresponding to the psychiatric wards were entered into the randomization software, and this software placed the psychiatric wards in either the intervention or control groups completely randomly and by coding. Participants were recruited from the same psychiatric wards across the two hospitals, and post‑randomization checks indicated no systematic imbalance in recruitment sources between the intervention and control groups.

To ensure allocation concealment, the randomization sequence generated for the participating psychiatric wards was kept by an independent researcher who was not involved in patient recruitment or intervention delivery. The allocation codes for the wards were inaccessible to the study team until the wards were assigned to their respective groups. The nurse responsible for participant enrollment was unaware of the ward allocation codes and therefore had no knowledge of which group each ward would be assigned to at the time of enrollment, ensuring concealment.

To minimize bias, data collection, questionnaire interpretation, and data entry and analysis were carried out by external research assistants and statisticians who were blinded to group assignments. In accordance with ethical principles, all patients were informed about the study’s focus on positive psychotherapy; however, participants were not informed about the specific intervention method (VR vs. face‑to‑face) used in the other group.

Materials and apparatus

A VR headset supplied by SHINECON G06A was used. This VR headset features a 185 x 133 x 100 mm screen and offers a wide field of view from 90° to 100°. It comes with adjustable IPD (interpupillary distance) ranging from 60 to 70 mm, HD resin aspherical lenses with a 40 mm diameter, and a fixed focal length of 37.5–46.5 mm. Additionally, it is compatible with smartphones running iOS, Android, or Windows Phone, accommodating screen sizes from 4.7 to 6 inches. We opted for this headset due to its notable features, including a wireless system, lightweight design (276 grams), comfort, and user-friendly interface. Furthermore, the headset is compatible with various mobile phone models and comes equipped with a controller (B01) to provide an immersive and captivating VR experience. The immersive 360° videos were designed based on the main concepts and fundamentals of positive psychotherapy by a team of experts, including professionals in the field of Unity game development, a C# programming language expert, and a 3D modeling specialist.

Intervention condition: VR-based positive psychotherapy

This intervention condition involved receiving three 30-minute VR sessions every other day. The sessions were designed based on Seligman’s positive psychotherapy [51], which is a suitable approach for enhancing well-being, fostering positive emotions, and promoting personal strengths in individuals. The number of sessions was adapted from the original protocol to enhance feasibility, consistent with previous evidence indicating that brief positive psychotherapy interventions can produce meaningful improvements in psychological well-being [25,52]. The sessions’ content was developed by a team of psychologists and psychiatric nurses under the supervision of the first author. They included 3D images, videos, sounds, and one or two exercises related to the contents of each session, which were prepared based on core positive psychotherapy concepts and practices, including kindness, gratitude, appreciation, identification of strengths, and goals. The intervention components were selected based on core mechanisms emphasized in positive psychotherapy, prioritizing elements that are theoretically central and well suited for experiential implementation within immersive VR environments. The content of the sessions is presented in Table 1.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Topics of sessions.

https://doi.org/10.1371/journal.pone.0354610.t001

Participants in the VR-based positive psychotherapy group wore a VR headset and initiated the session using a handheld controller. The VR program provided patients with immersive experiences in two distinct environments: a simulated consulting room designed to create a supportive and reflective atmosphere (see Fig 2), followed by a natural environment (e.g., forest, mountains, sky, and waterfall scenes), offering a tranquil and calming experience (see Fig 3).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Screenshots of the consulting room.

https://doi.org/10.1371/journal.pone.0354610.g002

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. Screenshots of the natural setting.

https://doi.org/10.1371/journal.pone.0354610.g003

Prior to the VR sessions, a trained psychiatric nurse (the second author) provided participants with brief explanations about the nature of the study, the intervention procedure, and the use of the VR headset and related equipment. Participants were instructed to focus on the session content and actively participate in the scheduled activities of the intervention. They were also encouraged to ask any questions to ensure familiarity with the equipment before the session began.

After the initial setup of the VR equipment, participants proceeded with the session independently. The psychiatric nurse remained in the room solely to supervise the procedure, ensure participant safety, and provide technical assistance if needed. Importantly, the nurse did not provide any psychological guidance or intervention during the sessions.

The VR intervention was designed as a guided, non‑interactive experience. The virtual environment did not include an avatar or any visual representation of the therapist. All intervention content and instructions were delivered exclusively through pre‑recorded audio narration embedded within the VR program, which guided participants through the activities and prompted reflection on personal experiences, such as recalling meaningful life events, identifying personal strengths, and expressing gratitude.

Participants could explore the virtual environment through natural head movements, allowing them to observe the surrounding immersive scene from different directions. A handheld controller was used exclusively to initiate the intervention and to transition between predefined audio‑visual parts within the virtual scenes.

Participants’ verbal responses were not processed by the VR program, and the virtual environment did not dynamically respond to user input. Instead, participants engaged in introspective reflection while following the audio instructions within the immersive visual scenes.

Control condition: face-to-face positive psychotherapy

The participants in the control group received three 30-minute face-to-face positive psychotherapy sessions every other day. The psychiatric nurse (the second author) provided these patients with consultations and exercises based on positive psychotherapy techniques, similar to what was provided for the intervention group. The in-person sessions were held in a private and quiet room in the psychiatric ward of the hospital.

Recruitment and procedure

Participant recruitment was conducted between January 20 and September 22, 2023. Outcome data were collected immediately before the first session (baseline) and after each of the three intervention sessions. No additional follow-up assessments were conducted after the final session.

At the outset, the research team obtained demographic and clinical data for 146 patients who had been diagnosed with unipolar MDD by a psychiatrist, based on DSM-V criteria. Following this, a psychiatric nurse (the third author) explained the aims of the study to the patients and inquired whether they were willing to participate in the screening process. She conducted screenings for the individuals who expressed interest until the target number of 86 participants was reached. All participants were informed that their participation was entirely voluntary, and they provided written informed consent.

Measures

In this study, sociodemographic and clinical characteristics were assessed using standardized forms developed by the research team. Data were collected on participants’ age, gender, marital status, educational level, and employment status, along with the duration of depressive symptoms and any family history of psychiatric disorders. The study utilized the following outcome measures:

Primary outcome

The primary outcome assessed in this study was emotional experience, given its significant influence on life satisfaction and suicidal ideation [10,11]. In line with the PERMA framework, emotional experience was prespecified as the primary outcome, whereas life satisfaction and suicidal ideation were considered secondary outcomes. Prespecifying a primary outcome helps clarify interpretation when multiple endpoints are examined [53].

To evaluate emotional experiences, the Positive and Negative Affect Schedule (PANAS) was employed. Originally developed by Watson et al. (1988), PANAS consists of 20 self-report items, equally distributed between positive and negative affect subscales. Participants rated their emotional states on a 5-point Likert scale ranging from ‘Very slightly or not at all’ to ‘Extremely,’ with scores for each subscale ranging from 10 to 50. Previous research by Watson and Clark (1988) demonstrated strong internal consistency for PANAS, with Cronbach’s alpha coefficients ranging from 0.86 to 0.90 for positive affect and 0.84 to 0.87 for negative affect across different time frames. The instrument also showed satisfactory test-retest reliability over an 8-week period, with correlation coefficients varying between 0.47 and 0.68 for positive affect and 0.39 to 0.71 for negative affect. Convergent validity was supported through its correlation with the Beck Depression Inventory, yielding coefficients between 0.36 and 0.58 over the ‘past few weeks’ [54]. In the Iranian context, Zargar et al. reported a Cronbach’s alpha of 0.80 for the negative affect subscale, affirming its reliability [55]. In the present study, PANAS exhibited acceptable internal consistency, with a Cronbach’s alpha of 0.78 for both the positive and negative affect subscales.

Secondary outcome

In this study, participants’ life satisfaction was assessed using the Satisfaction With Life Scale (SWLS), developed by Diener et al. (1985). This instrument consists of five items rated on a seven-point Likert scale, ranging from 1 (‘completely disagree’) to 7 (‘completely agree’), yielding a total score between 5 and 35. Score interpretations classify individuals into categories such as very high (30–35), high (25–29), average (20–24), slightly below average (15–19), dissatisfied (10–14), and strongly dissatisfied with life (5–9). Diener et al. (1985) evaluated the psychometric properties of SWLS among 176 undergraduate students, reporting a test-retest reliability of 0.82 over two months and a Cronbach’s alpha of 0.87. The scale demonstrated acceptable validity, with correlations of 0.50 and 0.37 with positive and negative affect scales, respectively [56]. In an Iranian context, Maroufizadeh et al. (2016) confirmed the reliability and validity of SWLS, with confirmatory factor analysis supporting a single-factor structure. Additionally, significant negative correlations were found between SWLS and the Hospital Anxiety and Depression Scale (HADS) for anxiety (r = −0.410) and depression (r = −0.434), indicating satisfactory convergent validity. The study also reported a Cronbach’s alpha of 0.887 for the scale [57]. In the current study, the internal consistency of SWLS, as measured by Cronbach’s alpha, was 0.91, demonstrating strong reliability.

In this study, the Brief Suicide Cognitions Scale (B-SCS) assessed patients’ suicidal ideation using a 6-item self-report scale. The B-SCS measures the suicidal belief system, encompassing enduring hopelessness related to self-perceived unlovability, emotional unbearability, and life problems perceived as unsolvable. The Likert-scaled items (1–5) focus on unlovability, unbearability, and unsolvability, yielding scores between 6 and 30. The B-SCS demonstrated strong reliability, validity, unidimensional factor structure, and predictive value in clinical settings [58]. The B-SCS showed robust internal consistency reliability (Cronbach’s alpha: 0.84–0.91) and strong test-retest reliability (0.84) in Rudd and Bryan’s study. In Iran, Batouei et al. validated the scale, revealing two significant subscales: unlovability and intolerability. Convergent validity was supported by correlations with Beck’s Suicide Ideation (0.57), while concurrent and divergent validity were indicated by correlations with Beck’s Hopelessness (0.62) and Beck’s Depression Inventory (−0.62), respectively [59]. The overall scale exhibited good internal consistency, with a reported Cronbach’s alpha of 0.85 in previous studies. In the present study, the reliability assessment indicated a Cronbach’s alpha of 0.82, confirming the acceptable internal consistency of the scale.

Feasibility, adherence, satisfaction, and sickness with VR-based positive psychotherapy

Feasibility in this study referred to the practicality of implementing the VR‑based positive psychotherapy intervention and conducting the study procedures within the clinical setting. Feasibility was assessed based on the proportion of eligible participants who enrolled in the study and completed it. A minimum of 70% adherence to the study protocol was considered indicative of acceptable feasibility [60]. Adherence to the intervention was determined by the number of modules and exercises completed by participants, as reported by themselves. Full adherence was defined as completion of all presented modules across the three sessions, along with providing feedback on the corresponding exercises.

In this study, two questionnaires, the Credibility/Expectancy Questionnaire (13) (Borkovec and Nau, 1972) [61,62], and the Virtual Reality Sickness Questionnaire (VRSQ) (Kim et al., 2018) [63], were used to assess satisfaction with the intervention and motion sickness in VR environments, respectively. The CEQ is a self-report measure comprising 5 items, with ratings on a Likert scale from 1 (indicating low/poor treatment expectation) to 10 (indicating high/favorable treatment expectation) [62]. In line with the present study, we employed a modified version of the CEQ [64] with five dimensions, including logical coherence (“How logical does this psychological program seem to you?”), overall satisfaction (“How satisfied are you with the psychological program?”), likelihood of recommending the program (“How confident would you be in recommending this psychological program to a friend who is experiencing a similar situation?”), perceived utility (“How useful do you think this program has been for you?”), and discomfort experienced (“How annoying or uncomfortable has this program been for you?”). Responses were assessed on a scale ranging from 0 (“not at all”) to 10 (“completely”). Both study groups completed this assessment at the conclusion of the third session. The VRSQ comprises nine items rated on a four-point Likert scale (0 = not at all, 3 = severe) and is divided into two subscales: the oculomotor system and disorientation. It assesses various symptoms such as fatigue, general discomfort, eye fatigue, difficulty concentrating, headache, motion sickness, blurred vision, feeling lightheaded, dizziness, and vertigo experienced during VR exposure [63]. This scale was completed by the patients in the intervention group after the third session. Both of these questionnaires have demonstrated good psychometric properties in previous research [61–63]. In this study, the VRSQ and the adapted CEQ demonstrated Cronbach’s alpha coefficients of 0.78 and 0.71, respectively.

Sample size

The sample size was calculated a priori based on the randomized controlled trial by Pietrowsky and Mikutta [65]. In that study, post‑treatment positive affect scores (PANAS) differed significantly between groups, with means of 26.89 (SD = 6.35) and 19.75 (SD = 5.92). The between‑group mean difference of 7.14, derived from these post‑treatment values, and the corresponding standard deviations were used as inputs for the present power analysis. Positive affect was therefore selected as the principal endpoint among our three outcomes (affect, life satisfaction, and suicidal ideation). Using MedCalc with α = 0.01 and power = 0.99, the required sample size was 37 participants per group; assuming 15% attrition, the target sample size was set at 43 per group (total N = 86).

Statistical analysis

Statistical analyses were conducted using IBM SPSS Statistics (Version 22.0; IBM Corp., Armonk, NY, USA). Descriptive statistics (means, standard deviations, and frequencies) were calculated to summarize demographic and clinical characteristics and to evaluate feasibility, adherence, participant satisfaction, and any adverse effects associated with the VR-based positive psychotherapy intervention. Baseline equivalence between the VR-based positive psychotherapy (VR-PPT) group and the face-to-face positive psychotherapy group was assessed using independent samples t-tests for continuous variables and chi-square (χ²) tests for categorical variables. All analyses were performed on a complete-case basis; participants with missing outcome data at the relevant time points were excluded from the corresponding analyses.

In this study, the independent variable was the mode of positive psychotherapy, categorized as VR-based positive psychotherapy (VR-PPT) or face-to-face positive psychotherapy. The dependent variables included positive affect (PANAS-PA), negative affect (PANAS-NA), life satisfaction (SWLS), and suicidal ideation (B-SCS), representing emotional experiences, life satisfaction, and suicidal ideation. These outcomes were measured at baseline and immediately before and after each intervention session.

Data normality was evaluated using the Kolmogorov–Smirnov and Shapiro–Wilk tests, and homogeneity of variances was examined using Levene’s test. The Kolmogorov–Smirnov test indicated no significant departures from normality for any variable. Although the Shapiro–Wilk test showed a mild deviation from normality for the life satisfaction variable, inspection of skewness and kurtosis values indicated that these departures were minimal. Considering the equal group sizes (n = 39) and strong empirical evidence that parametric procedures are robust against small violations of normality—particularly when such deviations are limited to slight skewness or kurtosis [66,67]— parametric analyses were considered appropriate for subsequent statistical procedures. Detailed normality results are provided in S1 File.

To examine intervention effects while accounting for multiple dependent variables, a 2 (Group: VR-PPT vs. face-to-face PPT) × 6 (Time: pre-and post-sessions 1, 2, and 3) mixed-model repeated-measures multivariate analysis of variance (MANOVA) was conducted. In this model, time was treated as the within-subject factor and group as the between-subject factor. Following significant multivariate effects, univariate repeated-measures ANOVAs were performed for each outcome variable to identify specific sources of change. The assumption of sphericity was evaluated using Mauchly’s test, and when violated, Greenhouse–Geisser corrections were applied. Post hoc comparisons were conducted to explore significant within-group changes across measurement points.

Effect sizes were reported to provide a more comprehensive evaluation of the intervention effects beyond statistical significance testing. Partial eta-squared (η²p) was used to estimate effect sizes for ANOVA analyses, and Cohen’s d was calculated for pairwise comparisons. Cohen’s conventional benchmarks (0.2 = small, 0.5 = medium, and 0.8 = large) were used to aid interpretation of effect magnitudes [68]. The magnitude of these effect sizes was considered when interpreting the findings. Statistical significance was set at p < 0.05.

Qualitative phase

Participants for the qualitative phase were selected using purposive sampling to ensure the inclusion of individuals with rich and direct experience of the VR-based positive psychotherapy program. In line with Patton’s (2014) recommendations, purposive sampling was employed to intentionally select “information-rich cases” capable of providing in-depth insight into the phenomenon under study rather than achieving statistical representativeness [69]. Accordingly, participants were eligible if they had: (1) completed the full VR intervention, and (2) expressed willingness to engage in an in depth interview and articulate their lived experiences. The second researcher (FAH) established telephonic communication with the participants, delivering comprehensive explanations regarding the purpose and scope of this study segment. Subsequently, for individuals expressing interest in study participation, she requested the completion and return of the provided informed consent through WhatsApp.

Each interview had a duration ranging from 30 to 45 minutes, and data were collected through semi-structured interviews. At the commencement of each interview, the researcher (the second author) posed some general and open-ended questions: “What is your opinion regarding the sessions and exercises delivered to you through the VR headset?” and “What changes did you perceive within yourself following these sessions?” Subsequently, the interview progressed to more detailed questions, such as “Why were you influenced by these trainings and exercises?.” If necessary, additional questions were directed towards the participant to elicit further clarification on her response and the subject mentioned. When applicable, probing questions like “Could you elaborate further?” “May you provide an example for better comprehension?” and “What does it mean?” were used.

Transcripts were reviewed several times immediately after each interview and then transcribed. Data analysis was conducted using the conventional content analysis approach proposed by Graneheim and Lundman (2009) [70]. This process involved several key stages, including identifying the unit of analysis, systematically coding the data, categorizing based on patterns of similarity and difference, condensing the data, and ultimately extracting overarching themes. Data collection continued until thematic saturation was reached, consistent with established qualitative standards [71]. Saturation occurred after 13 interviews, at which point no new codes, categories, or thematic variations were emerging from the data. Therefore, no further participants were enrolled in the qualitative phase. This sampling strategy and sample size are appropriate for exploratory qualitative studies embedded within mixed-methods designs.

To ensure the trustworthiness of the findings, Guba and Lincoln’s evaluative criteria were applied [72]. Credibility and confirmability were reinforced through prolonged engagement with participants over approximately seven months. To enhance credibility, initial coding underwent independent review by external research team members and three qualitative experts unaffiliated with the study. Additionally, member checking was conducted, wherein participants assessed the coded transcripts to ensure alignment with their lived experiences. Transferability was addressed by providing a detailed account of the research methods and findings. After completing the qualitative data analysis, the results of this phase were assessed and compared with those of the quantitative phase to determine homogeneity or heterogeneity.

Ethical considerations

The present study was conducted in accordance with the ethical guidelines outlined in the Declaration of Helsinki. Ethical approval was granted by the Ethics Committee of Shiraz (IR.SUMS.NUMIMG.REC.1401.059), and the study was officially registered in the Iranian Registry of Clinical Trials (IRCT20201001048893N7, registration date: 16 November 2022, https://irct.behdasht.gov.ir/trial/66424). Ethical considerations, including maintaining participant confidentiality and obtaining written informed consent, were rigorously upheld. Participants were fully informed about the study objectives, procedures, and their right to withdraw at any stage without consequence. Regarding data availability, relevant datasets can be accessed upon reasonable request by contacting the corresponding author via email.

Quantitative phase

The quantitative analyses examined changes in well‑being–related outcomes, including positive affect, negative affect, life satisfaction, and suicidal ideation, across the study period. Among the 86 patients deemed eligible for the study, 78 participants successfully completed the trial, with an equal distribution between the intervention (n = 39) and control (n = 39) groups. The primary analyses are based on participants with available outcome data at each time point (complete-case numbers). A total of 8 patients were excluded, comprising four individuals from each group. In the first post-treatment assessment of the intervention group, three patients didn’t complete the questionnaires, and one of them was discharged. Moreover, four patients were excluded from the control group in the first (two patients) and second (two patients) post-treatment assessments because of discharging [Refer to the CONSORT diagram (Fig 1) for a detailed overview of the participant flow].

The mean age of the participating patients was 35.38 ± 8.23 years. The participants had an average duration of depressive symptoms of 62.31 ± 54.29 months. The majority were female (91.03%), married (57.69%), and held an undergraduate degree (52.70%) (Table 2). Additionally, half of the participants (50%) reported a family history of psychiatric disorders. No significant differences were observed between the intervention and control groups in terms of demographic and clinical characteristics (Table 2).

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Sample characteristics (n = 78).

https://doi.org/10.1371/journal.pone.0354610.t002

The repeated measures MANOVA analysis demonstrated a statistically significant main effect of time (Wilks’ Lambda = 0.316, F(20,57) = 6.176, ɳ² = 0.684, p < 0.001), suggesting that the outcome variables underwent considerable changes over the study period. However, the main effect of the group did not reach statistical significance (Wilks’ Lambda = 0.918, F(4,73) = 1.633, ɳ² = 0.082, p = 0.175), implying that the overall differences between the intervention and control groups were not significant. Additionally, the interaction effect between time and group was not significant (Wilks’ Lambda = 0.658, F(20,57) = 1.483, ɳ² = 0.342, p = 0.124), indicating that the pattern of change over time was similar across both groups.

The results of the univariate analyses revealed a significant main effect of group on suicidal ideation (F[1,76] = 5.158, ɳ² = 0.064, p = 0.026), indicating a meaningful difference between the groups in terms of the marginal mean of suicidal ideation. However, no significant group effects were observed for positive affect (F[1,76] = 0.613, ɳ² = 0.008, p = 0.436), negative affect (F[1,76] = 0.208, ɳ² = 0.003, p = 0.650), or life satisfaction (F[1,76] = 0.636, ɳ² = 0.008, p = 0.428). Additionally, significant main effects of time were detected for positive affect (F[5,268.001] = 5.366, ɳ² = 0.066, p = 0.001), negative affect (F[5,258.389] = 22.153, ɳ² = 0.226, p < 0.001), life satisfaction (F[5,242.488] = 24.800, ɳ² = 0.246, p < 0.001), and suicidal ideation (F[5,271.372] = 17.947, ɳ² = 0.191, p < 0.001). These findings indicate that, independent of group differences, there were significant changes over time in positive and negative affect, life satisfaction, and suicidal ideation.

Because the intervention effect in longitudinal designs is reflected in the group × time interaction, this interaction was examined to determine whether the trajectories of change differed between the two groups. The interaction effect between group and time for positive affect, negative affect, life satisfaction, and suicidal ideation was not statistically significant (Table 3). This indicates that the patterns of change in these variables over time were similar across both groups, with no significant differences in their trajectories based on group membership.

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. One-way repeated measures MANOVAs, F-ratios, p values, and partial η².

https://doi.org/10.1371/journal.pone.0354610.t003

Pairwise comparisons between pre-treatment and post-treatment assessments indicated that both groups experienced improvements in positive affect, negative affect, life satisfaction, and suicidal ideation following the intervention. The means, standard deviations, and Cohen’s d effect sizes for the dependent variables across different measurement points are summarized in Table 4. The magnitude of the within-group effect sizes ranged from small to moderate across assessment points (d = 0.02–0.53), indicating generally modest changes in well-being-related outcomes in both groups over the course of the study.

Download:

• PNG
  larger image
• TIFF
  original image

Table 4. Repeated measures MANOVAs, means (SDs), and effect sizes (Cohen’s d).

https://doi.org/10.1371/journal.pone.0354610.t004

The means of positive and negative affect, life satisfaction, and suicidal ideation for both the experimental and control groups are illustrated in Fig 4.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. Estimated changes in negative and positive affects, satisfaction of life, and suicidal ideation in the pre-treatment and post-treatment measurements.

https://doi.org/10.1371/journal.pone.0354610.g004

Fig 4 illustrates the trajectories of the study variables across assessment points for the intervention and control groups. Overall, both groups exhibited comparable patterns of change over time. Notably, in the plot depicting negative affect, the plotted trajectories may appear non‑parallel at certain points; however, these visual deviations are inconsistent across sessions. Importantly, the magnitude of the between‑group differences in means at corresponding time points was minimal relative to the within‑group variability, and the within‑session effect sizes were similarly small in both groups. Under such conditions, fluctuations in the plotted means can produce visual deviations in the trajectories without indicating substantively different temporal trends. Accordingly, the pattern observed in Fig 4 is consistent with broadly similar trajectories of change in the two groups, in line with the non‑significant time × group interaction effects.

Individual trajectories and raw data values for all outcome measures are presented in the supplementary Spaghetti plots (S2–S5 Figs), which illustrate the variability in participant-level responses over time within each study group.

Out of the 43 participants who underwent VR-based positive psychotherapy, 39 (90.69%) successfully completed the post-assessments at T3. A total of three patients were excluded from the study for not finishing the questionnaires, and one was removed due to early discharge. Of the 39 patients who completed the post-assessments, 32 (82.05%) fully adhered to the intervention by completing all session modules and providing feedback on the exercises. The remaining participants showed partial adherence, with seven finishing the modules but not submitting feedback. In the control group, all 43 participants began the study, with 39 (90.69%) completing the post-assessments at T3. The other four were excluded due to early discharge.

In both the intervention and control groups, satisfaction levels were notably high, with item scores in the areas of logic, satisfaction, recommendation, and utility being 8 or higher on a 0–10 scale. Additionally, participants did not perceive the intervention as uncomfortable, as their ratings in the Uncomfortableness area were below 1 on the 0–10 scale (refer to Table 5). The t-test results revealed no significant differences between the mean scores of various satisfaction areas between the intervention and control groups (see Table 5).

Download:

• PNG
  larger image
• TIFF
  original image

Table 5. T-test statistics, means, and SDs: satisfaction with the intervention scale (adapted version of the Borkovec and Nau inventory).

https://doi.org/10.1371/journal.pone.0354610.t005

The VR sickness was assessed in the intervention group after completing the intervention sessions. The average total score of intervention participants on the VRSQ was 7.93 (SD = 11.45), indicating no noticeable or severe VR disease with the program among participants in the intervention group.

Qualitative phase

To deepen the understanding of the quantitative results, a qualitative phase was added to the study. In this phase, 13 individuals (9 females and 4 males) from the intervention group were interviewed using a semi-structured format via phone. The ages of the participants varied between 18 and 46 years. Table 6 presents the demographic details of the participants involved in this phase.

Download:

• PNG
  larger image
• TIFF
  original image

Table 6. Demographic characteristics of the participants.

https://doi.org/10.1371/journal.pone.0354610.t006

Two main themes emerged from the qualitative data, including “user-friendly program” and “efficient psychotherapy program.” The main themes, categories, and subcategories derived from the interviews are displayed in Table 7.

Download:

• PNG
  larger image
• TIFF
  original image

Table 7. Main themes, categories, and sub-categories in the qualitative stage.

https://doi.org/10.1371/journal.pone.0354610.t007

• User-friendly program

Two categories of this theme were “satisfactory design” and “client-centricity.” The participants believed that the design of the VR program was desirable and satisfactory. They also found the VR program had a constructive focus on them.

• Satisfactory design

The participants in the intervention group clearly mentioned their satisfaction with the VR program. The participants described the VR program as a high-quality and faultless program that motivated them to remain engaged with the VR sessions.

“In my opinion, the virtual reality program was excellent, and it had no issues or limitations; that’s why I truly used it” (P3).

“The software itself was complete in every aspect" (P2).

A participant pointed out that the technical aspects, including the excellent audio and visual quality, were satisfying, which contributed to her preference for using this program.

“The sound and image quality were good and satisfying, and in my opinion, this made me enjoy the program, leading to a greater inclination to use it” (P9).

They also believed that the program provided them with an opportunity to use a desirable, virtual environment, which improved their spirits and helped them feel healthy and positive. They recognized that delivering the psychotherapy program in a desirable nature-based virtual environment has a better effect on the learning of the participants. Two participants stated:

“Just by seeing the lush nature, water, and waterfalls and walking beside them, I thought that, like healthy individuals, I could enjoy the beautiful nature and have positive feelings. It was excellent. Utilizing this beautiful natural environment was very helpful in uplifting my spirits. The things mentioned in the program, such as expressing gratitude, were also wonderful in this atmosphere and had a positive impact on me” (P7).

• Client-centricity

The participants perceived a kind of empathy from the virtual psychotherapist while using the program. They felt that they were understood and that their suffering was validated.

“In the sessions we had, we felt a sense of empathy. For instance, when he said, “I understand; it is just like that; you’re right...” you felt that you were empathized with and that your sadness and suffering were validated, and for that reason, I think it had an effect on me and was a positive step in my treatment process" (P4).

Moreover, they believed that the program had a client-centric design, as the client’s main concerns and challenges were considered and addressed in the program.

“The teachings focused on important aspects that were actually my important challenges” (P5).

“In my opinion, this program was truly a very good option since it was designed for individuals with depression and to meet their needs” (P6).

The participants found that their privacy was preserved while using this program.

“In my view, this program makes the person more comfortable because someone may not feel comfortable talking face-to-face with a counselor about his issues or having others become aware of his problems" (P8).

• Efficient psychotherapy program

The participants in the intervention group clearly experienced the VR interventions as efficient. Categories associated with this main theme were “applicable psychotherapy method” and “psychological well-being developer.” Participants recognized the VR psychotherapy method as an applicable approach capable of enhancing their psychological well-being.

• Applicable psychotherapy method

The participants stated that the VR program is so applicable for them. In this way, they expressed that the program was so helpful and useful for improving their mood and keeping them feeling relaxed.

“The training was very helpful and practical for me, and it reduced my depression. I felt very relaxed” (P12).

The participants stated that the VR positive psychotherapy method is an adjunctive psychotherapy method that can improve their psychological condition.

“Your program, along with the medications I was taking, made me generally better. I mean, I felt much improved" (P13).

• Psychological well-being developer

The participants reported a development in their psychological well-being following the utilization of the VR positive psychology program. They highlighted enhancements in their spirit, positive feelings, and coping mechanisms. The participants also experienced a reduction in negativity post-intervention.

“The fact that I was going next to nature or using things related to gratitude was very good for improving my mood” (P3).

“It was very good training for me. I had a lot of negative feelings, thought about suicide, and felt guilty. After receiving these teachings, I learned to love myself, not to care about the past, and to think about positive things. I also didn’t let my mind focus on negative things and suicide anymore” (P11).

“Before using this program, I was very aggressive and irritable. I used to cry very quickly, but now I’ve found myself, and I’m not like that anymore. The training taught me that nothing is more important than one’s health and peace of mind” (P2).

Moreover, participants expressed an enhanced sense of self-connection. They remarked that they not only valued themselves but also nurtured self-love. They also stated that they accepted their past faults and guilt and had a profound capacity for self-forgiveness. Additionally, participants believed that they developed constructive cognitive self-awareness. They remarked on cultivating self-awareness of positive and negative thoughts, identifying self-empowerment, and reflecting on and remembering their blessings and strengths. Some participants stated that:

“As a result of this method, I was able to forgive myself, appreciate myself, and love myself. I could see the beauty and goodness in my life, recognize my capabilities, and be grateful for them. Moreover, I could focus on the positive aspects of my life. Essentially, the awareness I gained about my positive and negative emotions helped me prevent negative thoughts from coming back to me” (P1).

“I always had problems with my family, and I didn’t see my own faults and mistakes, and I didn’t accept them. Now, I have accepted my faults and mistakes, and I look at the circumstances realistically and optimistically" (P10).