The core elements of virtual forest exposure were classified into three categories: technology (device, medium, technical characteristics), forest (forest type, forest density, biodiversity, etc.), and participants (sociodemographic, natural experience, VR experience, etc.). The relationship between the three elements can be explained as follows: the virtual forest landscape is digitally processed and then exposed to participants through the virtual medium.

Devices (display systems) include a desktop display (2D), head-mounted display (HMD), and projection-based cavernous virtual system (CAVE). Among the included studies, 15 used 2D desktop displays, 45 used HMDs, and 3 used CAVEs [50,51,52]. The two-dimensional display systems were mainly laptop or TV screens, and the main HMD devices included the HTC Vive series (n = 16) and a series from Oculus (n = 13). HMD was the most commonly used device type (82.8%) in the included studies. VR with an HMD provides greater immersion and spatial presence than flat screen displays and is more cost-effective than CAVEs [53]. The media types included 2D photos or videos, computer-generated VR (CG-VR), and 360-degree VR based on real scenes (360-VR). Among the included studies, 360-VR was the most represented form of virtual media (51.7%). For 2D media, 14 studies included 2D photos/videos, of which 8 studies used only 2D photos/videos, and 6 studies were used for comparison with other virtual media forms or real forests. Nineteen studies included CG-VR: fourteen used only CG-VR, and five were used for comparisons with other forms of virtual media or real forests. In addition, the form of CG-VR varied, with three studies using an off-the-shelf VR game (Nature Trek VR) [54,55,56], and one study using digitally twinned real forests after they were scanned [22]. For 360-VR, 30 studies included 360-VR, wherein 23 studies used 360-VR only, and 7 studies were used for comparison with other forms of media or real forests. Most used 360-VR (video), and only one study used 360-VR (photo) [57]. Two-dimensional photos or videos are primarily captured using cameras or web sources. CG-VR is mainly generated using modeling software (e.g., Unity 2019, Unity Software Inc., San Francisco, CA, USA) or by directly using released games. Sources of 360-VR included ready-made videos shot by the research team using panoramic cameras or those downloaded from the web (Figure 3).

Several included studies considered differences between multiple exposure media, including comparisons between virtual and real forests and those between different virtual media (Table 1). Eight studies compared virtual and real forests: one compared real environments with 2D photos [61], two compared real environments with 360-VR [62,63], three compared real environments with CG-VR [22,64,65], one study compared real environments with 360-VR and CG-VR [66], and one study compared real environments with 2D videos and CG-VR [67]. An earlier study reported that although both 2D and real environments promote stress reduction, real environments rated significantly higher in terms of the degree of altered states of consciousness (ASC) and energy [61]. However, with advances in technology, virtual headgear has the potential to approach real nature and enhance the potential benefits [68]. A study comparing real forests and 360-VR forests found that both types of exposure to nature increased physiological arousal, benefited positive mood levels, and were restorative compared with indoor environments without nature [62]. One study reported similar benefits from real forests and 360-VR in terms of memory and executive function [63]. A study comparing a digital twin’s virtual forest with a real forest exhibited no significant difference in increased relaxation [22]. Although another study also demonstrated no significant difference between real and CG-VR forests in terms of both increased vigor and stress relief, the effect size was slightly stronger over time in physical conditions [64]. Interestingly, another study confirmed that CG-VR forests were just as restorative as physical forest environments; however, virtual forests appeared to be more fascinating and coherent [65]. Comparative real and virtual studies can help us understand the extent to which virtual forests replicate the benefits of real forests [69], which is important for assessing the potential of virtual forests [66]. However, research findings are not always consistent, with most studies reporting that virtual nature provides almost the same psychological and physiological benefits as real nature [69,70]. However, a few studies have propounded that virtual forests are more fascinating and compatible than real forests [65]. In conclusion, the extent to which virtual forests replicate the health benefits of real physical forests and whether this varies by virtual type and mode remains to be determined [70].

Seven studies compared the different types of virtual media. Although different virtual media play a role in reproducing nature, there may be differences among them. The main approach of relevant studies has transitioned from the use of still images to the use of dynamic media (e.g., video), interactive media (e.g., navigable), and finally to the use of fully immersive virtual media [71]. Four studies compared 2D and VR. Studies have demonstrated that highly immersive simulations can enable the perception of a higher quality of recovery than less immersive simulations, and that high levels of immersion can promote distance and perceptual fascination (360-VR vs. 2D laptop screens) [72]; however, another study reported that forest exposure in VR is ineffective in reducing stress when compared with 2D photographs [73]. Two studies described 360-VR and CG-VR comparisons. One of them reported that CG-VR forests were more emotionally restorative than 360-VR forests [66]; however, the other reported no difference between 360-VR and CG-VR forests in terms of generating a sense of presence, decreasing anxiety, or improving mood states. These results suggest that if the participants’ tasks are passive and do not require active exploration of the environment, then the use of 360-VR could be preferred because the development of 360° video is easier and cheaper than that required for computer-simulated environments [74]. Comparative multimedia studies are important to determine whether one (or more) should be recommended for recreation and treatment. Health authorities can compare the costs and benefits involved and provide guidance for different groups or situations [64]. However, studies between different media have also not reached consistent conclusions, and thus do not support the hypothesis that VR may be more effective and beneficial than traditional 2D media, which may be related to various factors such as video stability, luminance, and display content [70]. The most used medium in the included studies was 360-VR, and although it can provide a panoramic view to the experiencer and is more immersive than 2D video, its interactivity is limited, as it is prerecorded. CG-VR is highly interactive; however, this medium requires computer graphics skills for 3D modeling, and importantly, 360-VR is not an effective way to create a 3D model. VR, a non-interactive virtual landscape, does not fully mimic natural experiences, and longer experiments may also affect the participants’ perception of control and boredom [75].

Technical characteristics are unique features that differ from those of real forest environments. The technical characteristics of VR may be directly related to the effectiveness of the natural experience. The primary relevant terms included presence, realism, immersion, and motion sickness. Presence describes the subjective feeling of a person in a virtual environment, and is generated by the experience that is induced by immersive VR technology [29]. Motion sickness can be considered a specific type of visually induced motion sickness (VIMS) [76], which causes symptoms such as eyestrain, nausea, fatigue, headache, blurred vision, and postural instability during virtual experiences. Among the included studies, 39.7% addressed these technical characteristics, including motion sickness (n = 10), presence (n = 8), realism (n = 3), system usability (n = 2), immersion (n = 1), and tolerability (n = 1). However, most studies did not focus on or report on technical characteristics.

Studies have begun to focus on comparative studies on the differences in virtual technologies, including control, realism, and stability. In terms of the degree of control, one study explored its impact over one’s own behavior during nature experiences. The results indicated that active self-navigation had more positive health benefits than navigation by a researcher for a natural experience [55]. Regarding realism, one study compared the differences between CG-VR realism and demonstrated that more realistic VR environments evoked more positive emotional and serene responses and a greater sense of presence [67]. The results of another study indicated that realistic forest environments were more effective than dreamy forest environments in relieving psychological stress [77]. Another study reported that exposure to a CG virtual forest resulted in significantly higher cognitive performance, perceived restorativeness, positive emotions, sense of presence, lower perceived stress, and simulator disorders (motion sickness) than an abstract control group (cylinders, spheres, etc.) [60]. In terms of stability, the results of one study revealed that HIGH-stability 360-VR had a lower severity of motion sickness symptoms compared with LOW-stability video; however, no differences were found between the two in terms of presence, perceived environmental restorativeness, enjoyment, and HR.

First, several studies have reported that exposure to virtual forests has more positive health benefits than that to virtual urban environments, including physical and mental health [78], creative thinking [79], restorativeness [80,81], and nature connection [82]. In addition, virtual nature has more positive effects on stress reduction and cognitive improvement than indoor VR controls [17,58].

Second, in terms of comparative studies on virtual forest features and categories, the features addressed in the current study include forest type [83], natural light level [84], wildness [85], biodiversity [86], forest density [87], stimulation level [72], and perception of the video (perceived as live or recorded video) [88]. A study comparing the stress-reducing effects of seven different types of virtual forest resting environments (structure, wood, wood with a bench, wood with a platform and bench, platform with trees, waterfall with trees, and pool with plants) demonstrated that most natural environments do not have a very significant effect on stress reduction [83]. A comparative study on different levels of environmental brightness in a virtual forest demonstrated that exposure to moderately bright natural light significantly reduced stress in participants [84]. Another study reported that wild woods were described as more exciting than parkland and tended woodland [85]. A comparison of multiple levels of forest biodiversity revealed that low-biodiversity VR forests demonstrated the greatest improvement in most indicators of well-being [86]. A study reported that higher forest densities provide better physical and mental health benefits [87]. Comparisons of calming and stimulating nature scenes revealed that both triggered restorative outcomes through different psychological pathways, and stimulating nature scenes were more captivating and resulted in higher levels of presence [72]. One study even reported differences between naturally perceived live videos and recorded video media, suggesting that live videos had better attention recovery but the same stress recovery benefits [88]. This study echoes previous arguments that not only bottom-up (bottom-up) perceptual processes but also top-down (top-down) influences need to be considered when understanding how VR can be emotionally engaging.

In addition, the narrative of the forest environment story setting may have an impact on the effectiveness of the experience. In the design and optimization of virtual forest environments, the infusion of narrativity proves essential through the introduction of themes, role-playing dynamics, immersive audio elements, landscape variations, and interactive components [54,77,89]. Such meticulous design facilitates a cohesive and engaging storyline that increases participant engagement and fosters a more profound sense of presence in the virtual ecosystem, which in turn enhances the emotional experience [90].

Most participants in the included studies were healthy adults, with university students being the most abundant group. In terms of age and sex, two studies targeted older adults [91,92] and one study included middle-aged and older adults [93]. One study considered a sample of adolescents aged 14–19 years [56], one selected a sample of children [94], one included only males [50], and three selected pregnant women [59,95,96]. In terms of occupation, only one study included factory workers [97]. Subjects were also selected by mental health status, with one study selecting healthy adults with low natural connectedness (CN < 3.71) [82], one selecting older adults with cognitive or physical impairments [91], one selecting people who self-reported or were diagnosed by a doctor as suffering from stress or burnout syndrome [61], one selecting patients with generalized anxiety [52] (generalized anxiety disorder), and one selecting college students with mild-to-moderate anxiety and depression (stated anxiety and depression scale scores ≥40 and ≤60) [98]. Based on physical health status, one study chose breast and prostate cancer patients [99], one selected cancer patients in a treatment center [95], and one study included patients undergoing hemodialysis [100]. The number of included subjects ranged from 8 to 360. Although different types of participants were included in the reviewed studies, the most common type was healthy adults, particularly college students. Further large-scale trials are required to investigate whether similar effects are observed across a wider age range [40]. In addition to considering age, more research is needed to focus on specific populations, either occupationally or physically, who may be potential future users of virtual forest therapy and for whom it may be an effective way to engage in natural contact. Scholars have started focusing on the impact of individual characteristics. In addition to demographic characteristics such as age, sex, and education, the current study incorporates connections to nature and technology, as well as life experiences. The aspect of connection to nature includes nature connectedness [86,101,102], attitudes towards green spaces [103], landscape familiarity [75], nature experiences, and frequency of visiting nature [86,94,104]. The aspects of connection to technology, that is, familiarity with computers and VR, include VR experiences [50,56,67,93,100,105], gaming experiences [54], and knowledge of VR [106]. Life experiences include physical activity [94,104,107] and sleep quality [104].

The experience theme delves into the subjective and qualitative dimensions of individuals’ engagements with the virtual forest. It encompasses sensory stimulation, interaction, and dosage, providing insights into the experiential richness and perceptual dynamics of the virtual forest exposure. This category serves as a bridge between the core elements and the subsequent effects, as it explores how participants interact with and perceive the virtual environment.

In the reviewed studies, sensory stimulation was revealed to be assessed in 3 studies as visual only, 39 as audiovisual, 5 as audiovisual olfactory, and 1 as a supersensory system; however, 9 studies did not specify whether sensory stimulation other than visual stimulation was performed. This study used visual stimulation as the primary natural virtual exposure experience. Auditory studies have demonstrated an apparently significant interaction effect between vision and hearing in virtual natural environments, with natural sounds contributing to increased parasympathetic activity and more effective recovery after virtually induced stress [50]. Another study revealed that realistic sounds increase preference and realism [108]. Studies have shown that increased olfactory stimulation in VR natural environments can help reduce negative emotions and stated anxiety levels [89], and one study even suggested that olfactory stimulation may promote stress reduction more than visual stimulation [57]. A research protocol using VR and fogging phytofungicides (volatile organic compounds that are found in forested areas) to simulate forest immersion therapy was developed by a researcher to introduce positive changes in two patient groups: breast and prostate cancer patients [99]. Studies have demonstrated better recovery after multisensory exposure to 360-VR nature compared with visual-only experiences, and subjects perceive two identical nature scene landscape components significantly differently when auditory and olfactory stimuli are removed [86]. Studies have also begun to explore and propose a surreal multisensory nature experience called Nat(UR)e, where, in addition to audiovisual stimuli, subjects are exposed to natural odors and somatosensory stimuli, including wind, heat, and vibration [109]. The current research trend is to move from a predominantly visual experience to a multisensory experience. Scholars have suggested that to clarify the mechanisms by which multiple forest factors affect health, each sense needs to be stimulated separately through artificial climate laboratory experiments to determine the potential health benefits of forests and reveal the health-enhancing mechanisms of forest exposure [40]. Current virtual forest research mainly reflects the visual and auditory components of forests. The olfactory component can be added to a certain extent, but touch and taste are senses that cannot be easily simulated using current VR technology [22]. Regardless of technological improvements, it is likely that immersive virtual nature will never fully replicate the multisensory experience of real nature [29]. Regardless of the limited sensory stimulation, virtual forests may still produce similar effects to real forests, as a few recently published studies have shown [62].

The majority of subjects in the included studies were seated, with only a few studies involving other types of interactions. One study involved walking on a treadmill while wearing a VR mask [63], and another involved moderate-intensity aerobic exercise (cycling) in a CAVE virtual system [52]. A study that compared sitting and standing in a virtual environment reported an enhanced presence while standing [54]. Another study combined real physical activity with VR nature for children and found that physical activity followed by VR experience improved children’s attention levels [94]. Other studies simulated real forest walks in a VR forest using transient maneuvers [54,67,84]. Future research could consider incorporating more interactive designs into the virtual world to mimic real-world human behavior and allow subjects to become more immersed.

The duration of virtual forest exposure in the included studies ranged from 1 to 45 min. However, knowledge gaps remain regarding the effects of exposure duration on health benefits. A systematic review revealed that virtual nature interventions lasting more than 10 min produced more consistent effects than those of a shorter duration [110]. However, the optimal viewing duration may vary across media formats. A study on the effect of various durations of a 2D video of a forest revealed that viewing for 20 min produced significantly better relaxation than viewing for 5 or 10 min [111]. A study focusing on the natural dose of 360-VR reported that a 5 min exposure produced greater stress recovery benefits than 1 and 15 min [75]. However, another study did not find any significant effect of viewing duration on participants’ transient stress or mood [56]. In virtual environments, there may be a relationship between the exposure duration and motion sickness. Determining the optimal viewing duration may aid the practice of virtual forest therapy in the future.

Most studies have focused on the short-term effects of VR forest exposure, with limited research on the health benefits or potential adverse effects of long-term exposure (n = 9). Short-term exposure is generally defined as a single experience lasting from a few minutes to a few hours or a full day, whereas long-term exposure is defined as exposure lasting more than a day [112]. The causal role of natural exposure in improving short-term impacts is well known; however, its influence on long-term physical and mental health is not entirely clear [49]. The longest of the included studies lasted 12 weeks (one exposure per week) [97], and the shortest lasted two days (three exposures per day) [96]. One study reported that participating in a VR forest experience for three consecutive days resulted in sustained psychological improvements in older adults [92]. In another study, where three 10 min virtual nature exposures were conducted over two weeks, the interventions were restorative, resulting in significant and gradual increase in CN, which was strongly associated with the enjoyment of and motivation to participate in nature in the future [82]. However, a few studies did not find sustained health benefits in all cases. In one study, where subjects viewed forest videos for five consecutive days during COVID-19 isolation, only a short-term decrease in anxiety levels was observed, but not in the long-term, highlighting the limitations of virtual experiences [113]. Another three-week virtual nature intervention study reported no further improvement in the physiological health of patients undergoing hemodialysis after repeated VR exposure [100]. This result may be owing to repeated exposures to the same forest video, which led to boredom in subjects. Diverse forest landscapes may help reduce boredom and provide long-term benefits. Scholars emphasize that long-term studies are crucial for testing the efficacy of virtual forests, and that the optimal intervention exposure period for VR-based naturopathic treatments is still undetermined; thus, more reliable studies are needed.