Mindfulness, defined as the enhanced awareness of moment-to-moment experiences, where one intentionally attends to their thoughts, emotions, and bodily sensations in a nonjudgmental and accepting manner [1,2], has consistently been shown to provide wide-ranging psychological benefits across both general [3,4] and clinical populations [5,6]. Evidence from numerous studies indicates that mindfulness-based interventions are associated with a range of positive outcomes related to well-being, including higher levels of life satisfaction [7-9], vitality [7,10], optimism [11,12], pleasant affect [7,13,14], improved emotional processing [15,16], and emotional regulation [17-19]. Furthermore, these practices have been shown to effectively reduce depressive symptoms [7], stress [10,20,21], anxiety [22,23], and emotional reactivity [24-26]. These benefits arise not from suppressing or altering experiences, but from focusing on how individuals interpret their present-moment experiences [27,28]. When individuals perceive their thoughts and feelings as transient mental events that are impermanent in nature, they can reframe negative thoughts and decrease emotional reactivity, fostering a deeper sense of calm and well-being [29-31].

Beyond its mental health benefits, research has increasingly shown that mindfulness enhances cognitive performance [32-35] and is closely associated with improvements in executive functions, including inhibitory control, task-switching, and working memory [36-38]. The strong link between mindfulness and executive functions is theoretically well-grounded, as both rely on maintained attention and cognitive monitoring, indicating an overlap in their underlying cognitive mechanisms [39-41]. By fostering a relaxed mental state that promotes focused attention, mindfulness optimizes resource allocation and reduces mind-wandering, thereby improving cognitive performance [42,43].

In addition to cognitive benefits outlined above, mindfulness is closely linked to self-control. According to self-regulation theory [44], self-control operates as a limited resource that can be depleted under stressful and cognitively demanding conditions and replenished under restorative conditions. Cognitively demanding tasks require long-term executive control, such as maintaining attentional focus on goal-relevant information, suppressing distractions or impulses, and managing competing mental representations, and draw upon limited self-regulatory resources and temporarily reduce subsequent cognitive and emotional control capacity [45]. In contrast, restorative conditions are states or environments that facilitate the recovery of these depleted resources, including exposure to natural environments that promote attentional restoration, engagement in relaxation or mindfulness practices, or periods of rest that foster emotional rebalancing and cognitive recovery. When this resource is depleted, individuals may struggle with goal-directed behavior, impulse control, and regulation. A possible mechanism underlying this effect is that mindfulness enhances two fundamental self-control processes: (1) emotion regulation [17-19] and (2) attention regulation [33,46,47], both of which are crucial for maintaining self-regulatory capacity [44,48]. By enhancing these regulatory capacities, mindfulness provides a potential mechanism through which self-control resources can be replenished, thereby supporting long-term self-regulatory functioning.

Despite accumulating evidence on the positive outcomes of mindfulness on attentional and emotional restoration and self-control capacity, most findings are drawn from intensive programs, which typically span several weeks, requiring participants to engage in regular and often prolonged sessions under expert guidance [49]. Mindfulness-based stress reduction (MBSR) and mindfulness-based cognitive therapy are among the most pervasive and well-established interventions, with participants practicing mindfulness for up to 45 minutes daily and attending weekly group sessions that are 8 weeks long [2,50]. Specifically, 1 study demonstrated that MBSR is associated with reduced stress and enhanced psychological well-being [2], evidenced by improvements in emotional regulation, decreases in anxiety, and increases in life satisfaction among participants. Similar findings have been reported for other intensive mindfulness interventions. For example, a study found that participants who underwent a 1-month intensive mindfulness training reported significant reductions in anxiety and improvements in subjective well-being and self-compassion compared with a waitlist control group [51]. Additionally, another study found that novice meditators who participated in a 10-day intensive mindfulness training retreat demonstrated significant improvements in self-reported mindfulness, depressive symptoms, rumination, working memory, and maintained attention, relative to a comparison group who did not undergo any mindfulness training [32]. Generally, studies have highlighted the efficacy of intensive mindfulness interventions in improving well-being [52,53] and cognitive performance [54,55] compared with control conditions. Mindfulness fosters emotional regulation by increasing present-moment awareness and promoting acceptance of thoughts and feelings. According to the monitor and acceptance theory [56], mindfulness operates through two complementary mechanisms: (1) attention monitoring enhances awareness of moment-to-moment experiences, and (2) acceptance fosters a nonjudgmental and open stance toward these experiences. The interaction between these processes facilitates adaptive emotion regulation by allowing individuals to observe internal states without reacting impulsively.

Despite strong support for intensive mindfulness intervention, research on brief mindfulness has yielded mixed findings. On one hand, some studies report improvements in affect [57], reduced emotional reactivity, and enhanced attentional control [58,59]. Supporting this perspective, 1 study found that a single 15-minute mindfulness session significantly improved positive affect (PA) and reduced emotional reactivity in response to emotionally charged stimuli [60], highlighting the immediate benefits of mindfulness for emotional regulation. Similarly, another study showed that a single 15-minute brief mindfulness session could enhance attentional control by reducing mind-wandering during cognitive tasks [42], suggesting a notable impact on cognitive focus. While mindfulness has been associated with reduced emotional reactivity, it is important to note that reactivity itself is not inherently maladaptive. Emotional reactivity serves adaptive functions, which allow individuals to respond appropriately to meaningful positive or negative events. Excessive dampening of emotional responses, particularly to positive experiences, may reflect emotional blunting or reduced reward sensitivity, which can undermine well-being. Rather than eliminating reactivity, mindfulness promotes adaptive regulation, fostering flexibility to modulate emotions in a context-sensitive manner.

However, the outcomes of brief mindfulness practices are not always positive. Specifically, 1 study observed that a 15-minute mindfulness meditation did not significantly reduce negative affect (NA) relative to active control (guided progressive muscle relaxation training) and passive control (watching a TED Talks video) groups [61]. Likewise, another study found no improvement in working memory capacity following a 15-minute mindfulness breathing exercise in 2 high-powered studies [62]. These mixed findings highlight both the promise and limitations of brief mindfulness and underscore the need for approaches that can deepen engagement without imposing the burden of long-term practice [63,64]. One potential explanation for these inconsistencies is that brief mindfulness exercises may lack the long-term engagement and depth provided by intensive practices, such as MBSR and mindfulness-based cognitive therapy, which are essential for achieving optimal outcomes [49,65]. To address this limitation, recent technological advancements in virtual reality (VR) and mixed reality (MR) may offer promising solutions to enhance brief mindfulness practices.

VR and MR technologies provide avenues for enriching mindfulness practices by offering immersive environments that foster greater focus and engagement [66-68]. For example, one study found that VR provides an immersive environment that may reduce typical barriers to mindfulness practice by increasing engagement and reducing distraction [66]. Another study reported that integrating mindfulness with immersive VR shows promise for improving mood, attention, and engagement [67]. Finally, another study demonstrated that immersive VR conditions can produce greater mindfulness and stress-reduction effects than conventional mindfulness formats in some samples [68]. VR, in particular, provides controlled multisensory environments that simulate natural settings and promote focus [69,70]. Additionally, VR provides multisensory experiences that allow users to interact with virtual natural environments, helping them to disconnect from everyday stressors and deepen their mindfulness practice [71]. Meanwhile, MR, which integrates real and virtual elements, can further enhance mindfulness training by providing real-time feedback and increasing interactivity, which may further boost engagement [72,73]. Recent studies have explored the potential benefits of VR-based mindfulness, showing associations with reduced negative emotions [69,74-76], increased positive emotional states [66,77], and cognitive improvements in attention and working memory [46,78]. Yet, findings have not been entirely consistent, with some studies reporting significant effects on either PA or NA but not both [66,79].

Despite the growing body of research on VR and MR mindfulness interventions, important gaps remain, particularly concerning their effectiveness in brief applications and whether they provide emotional and cognitive benefits comparable with those of more intensive mindfulness practices. Most evidence supporting the cognitive benefits of VR-based mindfulness comes from studies involving multiple sessions, leaving uncertainty about whether shorter interventions yield similar positive outcomes [46,78]. For example, 2 separate studies reported having 8 sessions of the VR mindfulness intervention [46,78]. Moreover, many of these studies face methodological limitations, such as small sample sizes [66] and inadequate or inconsistent control groups [80,81], which constrain the robustness and generalizability of their findings. Research on MR-based mindfulness remains even more limited, despite its potential for enhanced interactivity [73]. These gaps are increasingly salient given recent technological developments, such as the Meta Quest 3 and Apple Vision Pro, which support both VR and MR modes. Emerging mindfulness applications also blend these modalities by embedding virtual cues within real environments or transitioning into full immersion when deeper focus is needed [82]. Thus, addressing these gaps through further empirical investigations is crucial to establishing the efficacy of brief VR and MR mindfulness interventions in enhancing emotional well-being and cognitive functioning and to determining their potential for scalable implementation across diverse real-world settings.

This study examined whether a brief mindfulness breathing exercise delivered via VR and MR can enhance emotional restoration, attentional functioning, and self-control capacity relative to a mind-wandering control condition. Using a within-subjects experimental design, participants completed both VR and MR mindfulness breathing exercises as well as a mind-wandering audio condition where participants were instructed to let their mind wander freely with no mindfulness advice provided, with a 1-week washout period between conditions to minimize carryover effects. We hypothesized that a brief mindfulness breathing intervention in VR and MR would lead to significant improvements in attentional and emotional restoration and self-control capacity compared with the mind-wandering control condition. Emotional and self-regulatory outcomes were assessed using standardized self-report measures, while working memory capacity was assessed using 2 well-established complex span tasks—the operation span (OSpan) task and the rotation span (RotSpan) task. These tasks were chosen for their strong validity and sensitivity to within-person changes in attentional functioning [83]. Repeated-measures ANOVA was used to examine condition-based differences across emotional, cognitive, and self-regulatory domains.

In this study, we selected PA and NA as immediate, subjective indicators of emotional well-being to capture participants’ affective states following mindfulness and VR or MR interventions. Although mindfulness has been shown to enhance several executive functions, including inhibitory control, task-switching, and working memory, we focused on working memory as it represents a central component of executive function and reflects the capacity to maintain goal-relevant information during mindful attention [36]. Previous research suggests that mindfulness training can improve working memory capacity by reducing mind-wandering and enhancing attentional stability, making it an appropriate measure for detecting short-term cognitive changes in a brief intervention [42]. We also conceptualized self-control within the limited resource framework of self-regulation, where momentary acts of regulation, attention control, or emotional management draw from a finite pool of control resources [84]. Because self-control sits at the intersection of executive control and emotional regulation [44], it serves as a meaningful link between our cognitive outcome (working memory) and affective outcomes (PA and NA). Hence, this study aimed to examine the effects of a brief (approximately 15 min) VR and MR mindfulness breathing exercise on attentional and emotional restoration and self-control capacity, compared with a mind-wandering control condition.

This study’s design and analysis plans were preregistered. The details of the preregistration are available on AsPredicted [85]. All preregistration documents and supplementary materials are publicly available on ResearchBox #3410 [86]. The study was reported in accordance with the APA Journal Article Reporting Standards (JARS) [87]. Frequentist analyses were conducted using IBM SPSS 29.0.1.0 [88], while Bayesian analyses were conducted using JASP version 0.17.3 [89] to examine evidence for null effects. The extent and pattern of missing data were examined across all primary outcome variables (PA, NA, self-control capacity, and working memory capacity) using SPSS Missing Value Analysis.

A total of 113 participants were recruited from a local university in Singapore. All participants were psychology undergraduates recruited using convenience sampling, and they received course credits for their participation. Each participant completed 2 counterbalanced sessions. Data from 11 participants were excluded from the analyses due to noncompliance with task instructions or insufficient immersion during either the experimental or control conditions, as indicated by experimenter observation. These participants appeared distracted or failed to engage consistently with the instructions. Specifically, 1 participant, who reported having dyslexia, was removed due to the potential impact of dyslexia on cognitive performance in the RotSpan task. Additionally, 1 participant encountered a technical error, and 9 participants failed to comply with study instructions. This resulted in a final analytic sample of 102 participants. Participant recruitment, eligibility screening, exclusions, assignment, and the final analyzed sample are summarized in a JARS-adapted flowchart in Figure 1. The demographic characteristics of participants are provided in Table 1.

The sample size of this study was determined using an a priori analysis in G*Power 3.1.9.7 [91] to ensure that the study was designed to meet the minimum sample size needed to detect an effect with 80% statistical power. Based on a medium effect size of f=0.25, an α level of .05, and at least 80% power (1–β), the analysis indicated that a minimum of 34 participants would be needed. Additionally, to ensure sufficient statistical power, a post hoc sensitivity power analysis was conducted using the same G*Power settings. This analysis indicated that with our final sample size (n=102), the study was adequately powered to detect the time×condition interaction effects specified by the study design as small as f=0.14 (equivalent to η_p²=0.019), which represents a small-to-medium effect size. No missing values were observed for any of the primary outcome variables.

All procedures were approved by the university’s Institutional Review Board (IRB-24‐119-A087-M2(1124)) and complied with its ethical guidelines. Participants were informed about the study aims, procedures, potential risks, and their right to withdraw at any time without penalty and provided written informed consent before the first session. To protect privacy and retain confidentiality, participants were identified only by unique identification codes. No personally identifiable information was stored with the research data. All analyses were conducted using deidentified datasets, with data access restricted to the research team. Participants received course credit in exchange for completing both study sessions. No images included in the manuscript or supplementary materials allow for the identification of individual participants. Informed consent was obtained from the individual depicted in Figure 2 for the use of their image in this publication.

This study used a 2 (condition: VR+MR mindfulness breathing vs mind-wandering)×2 (time: baseline vs postintervention) within-subjects experimental design. The within-subjects approach was chosen to control for interpersonal variability, thereby reducing errors associated with individual differences and enhancing the ability to detect true differences between conditions, ultimately increasing statistical power. All participants experienced both the experimental and control conditions. In the experimental condition, participants engaged in a brief VR and MR mindfulness breathing intervention that lasted approximately 15 minutes. In the control condition, they engaged in a mind-wandering audio that lasted approximately 15 minutes. The order of conditions was counterbalanced, with half of the participants completing the experimental condition first and the other half completing the control condition first. To mitigate potential carry-over effects, sessions were separated by a 1-week interval, which served as a washout period to allow any practice or residual effects from the first session to subside [92,93]. Participants underwent random assignment through Qualtrics to establish the sequence in which they experienced the 2 different conditions.

In the VR and MR mindfulness breathing condition, participants engaged in a mindfulness breathing exercise that lasted approximately 15 minutes using the Meta Quest 3 VR headset (Figure 2). Before the mindfulness breathing exercise, the experimenter instructed participants to prepare and explained how to use the controllers for 2 specific functions, which were used to navigate the virtual environments. Participants were informed that their VR screen would be shared on an external monitor (but not recorded) and reminded to follow the guided audio throughout the exercise. The VR and MR mindfulness breathing condition was administered through the Headspace XR app. Headspace XR offers both virtual and MR experiences, featuring mood-enhancing games, personalized guided meditations, and exercises designed to help users improve their mind-body connection through movement and breathing techniques. For this study, participants were tasked to engage in 3 mindfulness breathing activities within Headspace XR.

The first activity, “Take a Beat Portal,” adopts a technique known as “box breathing,” which takes place in an MR setting (Figure 5). In this activity, participants would first find an object or feature in the room that is shaped in a square. Then, they were instructed by the audio to draw a box-shaped virtual overlay that they would be using for the breathing exercise. Once the shape is drawn, participants complete three sets of four controlled breathing cycles. The guided audio included instructions, such as “Inhaling…feeling your chest and stomach expanding [...] Holding…feeling the breath in your body [...] Exhale… feeling the release of any tension [...] Holding… before the next breath.” This entire activity lasts approximately 5 minutes.

The second activity, “Firefly Treehouse,” is a breathing exercise accompanied by arm movements (Figure 5). This activity takes place in a virtual setting where participants are guided to take deep breaths, raising their arms while inhaling and lowering them during exhaling. The guided audio included directions, such as “Take in a deep breath, and as you inhale raise your hands up above your head [...] Hang there for a moment and as you exhale, bring your hands back to your side.” This entire activity lasts approximately 3 minutes.

The third activity, “Boxy Treehouse,” is a box breathing exercise similar to the first activity, with the only difference being that it is carried out in a VR setting (Figure 5). In this box breathing exercise, participants would similarly undergo 3 sets of 4 breathing exercises. The guided audio also included similar instructions as the first activity, such as “Inhaling…feeling your chest and stomach expanding [...] Holding…feeling the breath in your body [...] Exhale… feeling the release of any tension [...] Holding… before the next breath.” Throughout all 3 activities, participants were instructed to direct their attention to the present moment, feeling their breath, and being aware of each of their breaths. This entire activity lasts approximately 5 minutes.

In the mind-wandering condition, participants listened to a 15-minute audio track that allowed for free mind-wandering. Participants listened to the audio using headphones on the desktop. It included instructions, such as “Now we’re going to do an exercise for 15 minutes [...] Now simply think about whatever comes to mind, let your mind wander freely without thinking about anything in particular [...] Let your mind roam as it normally would [...] Allow your thoughts to wander wherever they may go [...] Go ahead and follow whatever thoughts that come to mind [...] Continue letting your mind wander, allowing your thoughts to wander wherever they may go.” Throughout the mind-wandering audio, there were no mindfulness instructions or interventions given.

The within-subject experiment (VR+MR mindfulness breathing vs mind-wandering) consisted of 2 sessions held 1 week apart in the laboratory. Data collection spanned 13 weeks during the semester, and analysis commenced only after the final session was completed. Participants were randomly assigned to either the VR and MR mindfulness breathing experimental condition or the mind-wandering control condition during their first session and assigned to the other condition in the second session, a week later. Throughout the study, the experimenter monitored participants to maintain data quality.

At the start of each session, participants received a link to the Qualtrics survey and were directed to the informed consent page. Participants used a desktop computer with headphones and a VR headset to complete the study. After providing informed consent, each participant generated a unique personal ID to be used across both sessions. Participants began each session by completing baseline measures, including the Circumplex Model of Affect, SMS-5, and OSpan. Following the baseline assessments, participants proceeded to complete activities in their assigned conditions.

After undergoing their respective conditions, a manipulation check was conducted to assess whether the participants were absorbed during the VR and MR mindfulness breathing intervention [42,98,99]. Participants rated the extent to which they felt absorbed in the present moment, focused on their breathing, and the physical sensations of their breathing. The 3 questions were asked using a 7-point Likert scale (1=Not at all absorbed, 7=Extremely absorbed), where higher scores correspond to greater levels of absorption. A manipulation check consisting of the same questions was conducted for the mind-wandering condition. Finally, they completed posttreatment measures, including the Circumplex Model of Affect, SMS-5, and RotSpan. Participants filled in their demographics at the end of the first session. At the end of the second session, participants were debriefed about the purpose of the study (Figure 6).

In line with our preregistration, we conducted a paired t test analysis, which revealed a significant difference in absorption between conditions, t₁₀₁=−16.29, Cohen d=−1.61, 95% CI −1.91 to −1.32; P<.001, indicating that participants reported greater absorption in the VR and MR mindfulness breathing condition (mean 15.46, SD 3.32) than in the mind-wandering control condition (mean 9.11, SD 4.07). Bayesian analysis provided extreme evidence in favor of the alternative hypothesis, BF₁₀>100 [100], indicating a substantial difference in absorption levels between conditions.

Following our preregistered analytic plan, differences in working memory scores between conditions via a 2 (condition: VR+MR mindfulness vs mind-wandering)×2 (time: baseline vs posttreatment) were tested using repeated-measures ANOVA. A significant main effect of time on working memory was observed (F_1,101=149.34, η_p²=0.597; P<.001). However, there was no significant main effect of condition (F_1,101=0.343, η_p²=0.003; P=.56). Similarly, the interaction effect between condition and time on working memory capacity was not significant (F_1,101=1.366, η_p²=0.013; P=.25), and working memory scores differed similarly between baseline (mean 0.90, SD 0.13) and post treatment (mean 0.74, SD 0.18) in the VR and MR mindfulness breathing condition, and between baseline (mean 0.88, SD 0.15) and post treatment (mean 0.74, SD 0.17) in the mind-wandering control condition (Figure 7). This indicates that the difference in working memory performance between baseline and post treatment did not differ between the VR and MR mindfulness breathing condition and the mind-wandering control condition. Bayesian analysis further provided moderate evidence in support of the null hypothesis for both the main effect of condition and the time×condition interaction (BF₀₁=7.42 and BF₀₁=5.55, respectively), suggesting that VR and MR mindfulness breathing did not produce measurable improvements in working memory capacity compared to the control condition.

Consistent with our preregistered analytic plan, repeated-measures ANOVA was conducted to examine whether levels of PA were influenced by time and condition. First, no significant main effect of time on PA was observed after Bonferroni correction (F_1,101=5.367, η_p²=0.050; P=.02). This indicates that PA did not change significantly across time, regardless of the condition. However, a significant main effect of the condition on PA was found, F_1,101=13.278, η_p²=0.116; P<.001, indicating that the VR and MR mindfulness breathing condition showed significantly higher levels of PA (mean 26.47, SD 6.73) than the mind-wandering control condition (mean 24.20, SD 5.95). A Bonferroni correction was applied to account for multiple comparisons. With a corrected significance threshold of P<.0167 (ie, .05/3) for 3 comparisons, the main effect remained significant.

Most importantly, the interaction effect between condition and time on PA was significant, F_1,101=58.30, η_p²=0.366; P<.001, suggesting that the changes in PA scores over time differed significantly between the VR and MR mindfulness breathing condition and the mind-wandering control condition. Specifically, PA scores increased from baseline (mean 24.59, SD 6.74) to posttreatment (mean 28.35, SD 7.63) in the VR and MR mindfulness breathing condition, whereas scores in the mind-wandering control condition declined from baseline (mean 25.03, SD 6.24) to posttreatment (mean 23.36, SD 7.25), reflecting a statistically significant difference between conditions (Figure 7). This effect remained significant under the Bonferroni-corrected threshold (adjusted P<.0167). Bayesian analysis further provided extreme evidence supporting the alternative hypothesis (BF₁₀>1000, BF₀₁<0.001). The 95% credible intervals for the cell means indicated that PA decreased in the control condition from 25.03 (23.80-26.26) at baseline to 23.36 (21.94-24.79) post treatment, whereas it increased in the VR and MR mindfulness condition from 24.59 (23.26-25.91) to 28.35 (26.86-29.85).

A similar analytical approach was used to examine changes in NA. A significant main effect of time on NA was observed, F_1,101=62.619, η_p²=0.383; P<.001, suggesting that NA changed significantly from baseline to post treatment. While there was no main effect of the condition on NA, F_1,101=1.384, η_p²=0.014; P=.24, the interaction effect between condition and time on NA was significant (F_1,101=39.087, η_p²=0.279; P<.001). The results showed that the change in NA scores over time was significantly different between the VR and MR mindfulness breathing condition and the mind-wandering control condition. Specifically, NA scores declined from baseline (mean 14.15, SD 4.84) to post treatment (mean 10.58, SD 2.74) in the VR and MR mindfulness breathing condition, whereas the decline was less pronounced in the mind-wandering control condition from baseline (mean 13.15, SD 5.12) to post treatment (mean 12.36, SD 4.27). This effect remained statistically significant under the Bonferroni-corrected threshold (adjusted P<.0167). Bayesian analysis also provided extreme evidence in favor of the alternative hypothesis (BF₁₀>1000, BF₀₁<0.001). The 95% credible intervals indicated that NA decreased modestly in the control condition, from 13.15 (12.14-14.15) at baseline to 12.36 (11.52-13.20) post treatment, but showed a substantially larger decrease in the VR and MR mindfulness condition from 14.15 (13.20-15.10) to 10.58 (10.04-11.12).

Similar analyses were conducted to examine changes in self-control capacity. A significant main effect of time was observed, F_1,101=15.480, η_p²=0.133; P<.001, indicating that self-control capacity changed significantly from baseline to post treatment. Moreover, a significant main effect of condition was found, F_1,101=24.910, η_p²=0.198; P<.001, and remained statistically significant after the Bonferroni correction. Consistently, the interaction effect between condition and time on self-control capacity was also statistically significant (F_1,101=28.348, η_p²=0.219; P<.001). The results showed that self-control capacity scores increased from baseline (mean 22.09, SD 4.60) to post treatment (mean 25.25, SD 4.42) in the VR and MR mindfulness breathing condition, whereas scores in the mind-wandering control condition decreased from baseline (mean 21.77, SD 5.14) to post treatment (mean 21.31, SD 5.46). Bayesian analysis further provided extreme evidence in favor of the alternative hypothesis (BF₁₀>1000, BF₀₁<0.001). Collectively, these findings suggest that exposure to VR and MR mindfulness breathing was associated with an improvement in self-control capacity.