There is an urgent need to design and implement effective interventions to decrease the incidence of mental health disorders in real-world settings. The most effective primary intervention for treating patients with MDD and GAD in primary care is pharmacological (e.g., antidepressant) therapy (Connolly and Thase, 2011; Dwyer and Bloch, 2019; Montano et al., 2023; Sharma et al., 2017). This approach has been found to be clinically effective in reducing the symptoms of mental health disorders of patients (Alanko et al., 2022; Leichsenring et al., 2016; Meadows et al., 2014). Despite the beneficial effects of pharmacological therapy, a growing body of literature has reported risk of harms related to the use of commonly used pharmacological treatments and has identified mild adverse side effects such as sleep disturbance, sexual dysfunction, increased blood pressure, high dropout rates, and low remission rates (Connolly and Thase, 2011; Dwyer and Bloch, 2019; Montano et al., 2023; Sharma et al., 2017). Thus, primary care patients and clinicians have stressed the importance of non-pharmacological treatments that decrease MDD episodes such as complementary and alternative medicine, exercise, or a combination of non-pharmacological treatment and antidepressant medication (Alanko et al., 2022; Leichsenring et al., 2016; Meadows et al., 2014).

Mindfulness-based cognitive therapy (MBCT) is a non-pharmacological approach to ameliorating negative emotions, GAD and MDD symptomology, and related comorbidities (Mennin et al., 2015; Saeed et al., 2019). Emotional regulation was found in systemic biomarkers as one of the neuropsychological benefits of the MBCT (Tolahunase et al., 2018). The improvements in emotional regulation realized through MBCT contribute to a reduction in depressive symptoms (Alanko et al., 2022; Leichsenring et al., 2016; Meadows et al., 2014), cortisol levels (Zhang et al., 2023), oxidative stress (Liu et al., 2015), and an increase in metabolism (Henje Blom et al., 2015), serum Brain-Derived Neurotrophic Factor (BDNF) (Autry and Monteggia, 2012) and neuroplasticity affected by MDD (Eyre and Baune, 2012).

The effect of MBCT (i.e., meditation) in preventing the relapse and exacerbation of depression has been described in longitudinal clinical trials (Meadows et al., 2014; Bédard et al., 2014; Edenfield and Saeed, 2012). Our recent pilot trial found that the MBCT was effective in reducing MDD, GDD and related mood disorders, and that the effects were maintained at a nine-month follow-up period (Renna et al., 2018). Two years of longitudinal study identified that the MBCT was effective in reducing depression and its recurrence in a follow-up assessment (Edenfield and Saeed, 2012). Moreover, MBCT reduced depression in patients with post-traumatic stress disorder (PTSD) through a three-month follow-up assessment (Bédard et al., 2014).

Digital health technology is a novel approach to care for mental health conditions whose feasibility and efficacy have been tested in various clinical settings (Jacobson et al., 2019; Naslund et al., 2017; Patel and Butte, 2020). Digital health encompasses a wide range of technology-based interventions (i.e., complementary treatments), precision medicine, and telehealth care that have been successfully applied to mental healthcare for individuals with MDD, GAD, and related comorbidities (Anderson et al., 2022; Arafat et al., 2021; Torous et al., 2021; Mandryk et al., 2021; Rykov et al., 2021). These technologies take on the role of predicting mental health status based on objective data over time and providing a cost-effective standardized service (Anderson et al., 2022; Arafat et al., 2021; Torous et al., 2021). Virtual reality (VR) technology is an example of a cost-effective solution for providing complementary mental health care in a consistent, standardized, precise manner (Park et al., 2019; Chirico and Gaggioli, 2019). The integration of VR technology with MCBT programs is designed to increase the efficacy of therapy for emotional regulation in MDD and GAD patients and decrease the treatment dropout rate (Colombo et al., 2021). VR interventions yield superior outcomes in managing symptoms as compared to traditional mental health care for MDD and GAD (Gorini and Riva, 2008). Recent clinical trials have provided evidence that meditation facilitated by VR technology significantly enhances depression and anxiety care (McIntyre et al., 2023; Paul et al., 2020; Paul et al., 2022). VR treatment has proven its efficacy in addressing a wide range of mental health disorders. Exposure to positive emotion in a VR meditation program was associated with a significant reduction in the agitation, depression, anxiety, and negative emotion of individuals with MDD (Chen et al., 2020; Ioannou et al., 2020).

There have, however, been several shortfalls noted in previous studies. First, although the health benefits of CBT and MBCT as non-pharmacological treatments have been reported, little research has been conducted to investigate the benefits of immersive VR for individuals with MDD and GAD. A growing body of studies found that an IVRM program can address practical limitations to participation by: (a) accommodating individualization by allowing participants to choose from a variety of nature-based themes to navigate and interact with in over 60 different natural settings in more than 500 locations, (b) reducing physical and structural barriers, a benefit inherent to VR, by enabling participants to access nature-based settings at their preferred times and locations, and (c) addressing individual challenges associated with physical limitations (e.g., mobility, risk of falling, and pain), financial and weather-related issues, and the need for specially modified equipment (Park et al., 2019; Chirico and Gaggioli, 2019; Glegg and Levac, 2018). Even though previous literature has reported that MBCT integrated with technology is a feasible and cost-effective non-pharmacological treatment, there is, currently, a dearth of research investigating its clinical effects in patients with MDD and GAD. Second, objective measurement is required to improve the reliability and validity of psychometric assessments. The information collected from patients with MDD and GAD in previous reviews and case studies was limited due to unreliable feedback (Kandasamy and Campbell, 2022). It must also be noted that most clinical trials investigated the relationship between intervention engagement and health outcomes. However, further investigation is required to gain an understanding of the benefits of the intervention on health outcomes.

The goal of this study was to investigate the longitudinal relationship between improved emotional regulation resulting from participation in Immersive VR Meditation (IVRM) and MDD and GAD as monitored by electrocardiogram. This study will provide evidence for improved emotional regulation function related to the use of an IVRM program on the anxiety and depression of patients with MDD and GAD.

The sample in this study consisted of inpatients in the behavioral healthcare unit of a community hospital located in Bloomington, IN from January to April 2023. This specialized unit delivers comprehensive residential psychiatric care to individuals afflicted by a spectrum of mental health challenges including, but not limited to, schizophrenia, bipolar disorder, depression, and dissociative disorders and provides treatment for behavioral manifestations including substance abuse, contemplation of self-harm, and suicidal ideation for up to 14 days.

The inclusion criteria for participants in this study were: (a) a clinical diagnosis of MDD, and (b) confirmation of a diagnosis of GAD by the attending unit psychiatrists. Potential participants manifesting symptoms of schizophrenia, dissociative disorders, psychosis, and impairments in visual and auditory ability were omitted from the study sample.

This study was a longitudinally designed, single-arm clinical trial in which the intervention was conducted three times a week for 10 weeks. We recruited participants based on referrals from physician progress notes and initial entrance exams. Research assistants individually contacted the patients diagnosed with MDD and/or GDD and obtained informed consent from the legal representatives of all participants before their engagement in the program. However, not every participant received the same number of interventions for two reasons: (a) even though the intervention period was 10 weeks, some participants were not able to attend every session due to their daily mental health condition and medication schedule, and (b) the physicians in charge of the emergency unit discharged some of the patient participants earlier than was planned. The discharge criteria were confidential, and the research team was not allowed to access the treatment schedule in the unit. Thus, some participants ceased participation in the intervention before the conclusion of the planned course of the intervention.

Of the initial group of 36 participants, each person, on average, engaged in around 5.1 sessions. However, 11 participants left the study without undergoing the exit assessment when they were discharged from the hospital. This left 25 participants who took part in an average of 2.7 intervention sessions which resulted in a total of 68 observations.

This study was implemented following the ethical guidelines of the Institutional Review Board (IRB: IU#17808). We solicited and received written informed consent from the participating patients themselves or their legal representatives holding a power of attorney prior to the initiation of the data collection and treatment phases. To ensure comprehensive understanding of the study objectives and procedures by participants in the intervention group, detailed information encompassing the title, objectives, duration, type, and the nature of study participation was provided.

This study employed Guided Meditation VR (Cubicle, Ninjas, United States) as a technology-based innovative approach alongside traditional mindfulness programs. The Oculus Quest 2 digital headset developed by Meta was used to provide the Guided Meditation VR. The most important aspect of IVRM that distinguishes it from traditional meditation is the provision of a personalized meditation experience. Participants were able to explore natural scenery such as a green meadow, waterfall, savannah, and beach and choose a variety of nature sounds (e.g., birds chirping). Second, the IVRM provided a wide range of guided meditation programs from which participants could choose a specifically designed meditation program to release stress, assist sleep patterns, and reduce depression (Ma et al., 2023; Shaw et al., 2011). The 30-min IVRM sessions were provided according to the preference and requirements of each participant.

HeartMath (i.e., electrocardiogram) was used to measure the changes in emotional regulation related to IVRM participation. HeartMath is a biofeedback monitoring system that measures heart rhythm changes and coherence levels between sympathetic and parasympathetic activities in the autonomic nervous system (ANS) (Raghavendra and Telles, 2012). This biofeedback data allows researchers to objectively measure coherence level and related achievement scores that are associated with levels of depression and anxiety (Drageset et al., 2012; Edwards, 2016; Minen et al., 2021). The Coherence Achievement Score in HeartMath indicates changes in the level of depression and anxiety after clinical interventions.

Prior to the main analysis, we explored the demographic characteristics of study participants and conducted a descriptive analysis of the pre-post CAS gap, depression, and anxiety. The main analysis longitudinally investigated the relationship between CAS gap parameters, depression, anxiety, and covariates. The Generalized Estimated Equation (GEE) was employed to investigate the longitudinal relationship between the CAS gap (i.e., parameter), depression, and anxiety after the IVRM intervention. Age, sex, marital status, and educational level were entered as covariates into the GEE model.

The reason for using GEE is due the inability of study participants to attend every session that was provided three times a week for 10 weeks. This irregular participation caused fractional data in which the number of interventions differs between participants and were censored occasionally during the data collection. Notably, the fractional data (e.g., censored data) were inherently heteroskedastic, meaning that the variance of the outcome is tied to the values of the predictors. This results in inflated standard error estimates in both linear and logistic regression models that results in reduced statistical power for detecting a treatment effect. To address this concern (e.g., multicollinearity), applying GEE is the most appropriate approach that could be used to analyze the fractional data in longitudinal studies such as this project (Papke and Wooldridge, 2008).

The longitudinal relationships between the parameter (CAS differences), anxiety, and depression were investigated, and a 95% Confidence Interval (CI) was reported. In summary, the GEE model computed the changes in depression and anxiety as an increasing parameter (i.e., slope coefficient). All statistical analyses were conducted using SPSS version 28.0. The model equation used was:

Outcome Variable = Constant + Covariates + CAS differences

Depression, Anxiety =  β 0 + {(Age* β 1 ) + (Gender* β 2 ) + (Marital status* β 3 ) + (Educational level* β 4 ) + (CAS differences* β 5 )}

Participant ages ranged from 18 to 71, with a Mean (M) of 42.51 and a Standard Deviation (SD) of 19.60. The sample consisted of 44.0% males (n = 11) and 56.0% females (n = 14) (Table 1). Two thirds of the participants were divorced (66.7%, n = 16) and on average had a high school educational level (58.4%, n = 14). Study participant descriptive statistics are summarized in Table 2. The average number of completed interventions per participant was 2.72 (SD = 1.20).

The CAS gap score at the exit assessment (M = 89.94, SD = 35.27) increased from the CAS gap at baseline (M = 54.69, SD = 22.95), and depression (M = 29.16, SD = 3.72) and anxiety (M = 23.52, SD = 2.14) scores increased at the exit assessment compared to baseline assessments (depression: M = 26.16, SD = 2.56; anxiety: M = 20.33, SD = 4.08). The effect size (i.e., Cohen’s d) is presented in Table 2. The CAS difference was 22.25 (95% CI: 1.22–1.93), with depression at 6.86 (95% CI: −1.11 to −0.23), and anxiety at 4.03 (95% CI: −1.21 – 0.31).

Table 3 shows the relationship between parameter estimate and depression. The unstandardized beta (B) represents the slope coefficient of depression increasing with CAS gap. The parameter ranged from 2 to 84 and the slope coefficient (B) was between −5.05 to 7.33. Depression significantly decreased with increasing CAS gap and showed a linear pattern from a parameter of 2 to a parameter of 84. The highest slope coefficient was 7.33 (SD = 0.53) at the parameter of 84 (95% CI: 129.21–394.01). The lowest B value was −5.05 (SD = 0.90) at the parameter of 5 (95% CI: 0.00–8.35). The B value of age was negative (B = −0.06, SD = 0.02, 95% CI: 0.90–0.98), indicating that older individuals had higher levels of depression. Females were found to have lower depression scores than males (B = 0.72, SD = 0.41, 95% CI: 0.91–4.66) and highly educated participants were found to have lower depression scores than individuals with lower education levels (B = 0.65, SD = 0.19, 95% CI: 1.30–2.80).

Table 4 displays the relationship between parameter estimate and anxiety. The parameter ranged from 2 to 84, and the B value ranged from 5.60 to −0.93. The decreasing slope of anxiety was linear with increasing CAS gap scores. The highest B value was 5.60 (SD = 2.50) at a parameter of 84 (95% CI: 2.03–22001.12). The lowest coefficient value was −0.93 (SD = 2.88) at a parameter of 5 (95% CI: 0.08–19.82). The coefficient of age was −0.04 (SD = 0.06, 95% CI: 0.85–1.08), indicating that older participants had higher anxiety levels than their younger counterparts. The sex coefficient was −3.38 (SD = 1.66, 95% CI: 0.01–0.33), indicating that male participants had lower anxiety levels than females.