Can Even a Small Amount of Greenery Be Helpful in Reducing Stress? A Systematic Review

A positive experience of nature triggers beneficial mental and physical responses. Today, we live in a rapidly urbanizing world where access to nature is often limited. Against this backdrop, this systematic review investigated studies on the effectiveness of small-scale greenery for stress reduction. We searched EMBASE, Cochrane, Web of Science, Scopus, PubMed, and Science Direct, searching databases from inception to April 2022. Studies were screened against predetermined criteria, and the risk of bias was assessed using the Cochrane Handbook for Systematic Reviews of Interventions for RCTs and The Risk of Bias in Non-Randomized Studies of Interventions (ROBINS-I) tool. Of the 2500 records identified, we screened 1817 citations for eligibility, which included 13 RCT studies and 6 non-RCT studies. The studies were conducted in eight different countries. The study populations included office workers, students, senior citizens, and patients with specific diseases. Research has mainly focused on indoor greening, with relatively little research on small-scale outdoor greening. All included studies assessed the impact of the intervention on various stress reduction-related outcomes, with the most common stress measures being blood pressure and the State Trait Anxiety Inventory (STAI). Various beneficial effects of the interventions on human health were reported in all 19 studies, 15 of which reported positive effects on stress reduction. All included studies were at high risk of bias. It is recommended that future studies in this area take appropriate measures to reduce bias and improve quality in order to build a strong evidence-based medical foundation. According to our findings, even very small-scale greening, including indoor green walls and potted plants, may provide effective help for stress relief. Understanding the physiological and psychological benefits of small-scale greenery can help better provide more opportunities for urban residents to engage with nature in the context of dense urban trends, as well as provide some reference for urban design planning.


Introduction
Human health and well-being have always been affected by the quality of the environment in which people live [1]. With urbanization, a large percentage of the population lives in stress-filled environments [2]. Previous research has found that people who grow up in cities have higher rates of psychiatric disorders, anxiety disorders, and depression than those who grow up in rural areas [3]. These disorders are often associated with or triggered by high levels of stress [4]. Symptoms of stress may include decreased memory and concentration, insomnia, increased heart rate, headaches, and muscle aches and pains [5]. The consequences of these symptoms may include decreased functioning socially and at work. Chronic stress can suppress the immune system and trigger cardiovascular disease, stroke, depression, asthma, and other serious health problems [6].
Numerous studies have shown that positive experiences in nature trigger beneficial psychological and physiological responses, such as lower blood pressure and heart rate, reduced muscle tone, better concentration, lower stress hormone levels, and improved geographical regions were eligible, but only references written in English were included. Detailed search strategies are shown in the Supplementary Materials.

Inclusion and Exclusion Criteria
Inclusion criteria were defined according to PICOS (Population, Intervention, Comparison, Outcome, and Study Type) policies. For the population (P), namely the participants under consideration, any adult was eligible regardless of their physical or mental health status. For intervention (I), the experimental group was a small-scale greening. Small-scale greening was defined as green facilities less than 0.4 ha in size, including outdoor greening, such as green roofs, pocket parks, private gardens, and street trees, and also small-scale green features, such as indoor green walls, decorative planters, and greenery (virtual environments were also considered) [18,[30][31][32]. For (C), no greening or any other comparative intervention within the environment. The primary outcome (O) was stress reduction. This should be a quantitative measure. The measure must refer to current physiological stress responses, as well as self-reported psychological indicators (e.g., mood, anxiety, distress, perceived stress, recovery, attention, or cognitive functioning). (S) Inclusion criteria were limited to experimental clinical studies with interventional factors. Subjects were categorized into randomized controlled and non-randomized controlled studies based on whether or not they were randomized, and observational studies without intervention factors were not included. As shown in Table 1. Table 1. Inclusion and exclusion criteria.

Inclusion Criteria: Exclusion Criteria:
Empirical studies. Studies that do not look at empirical evidence.
Studies examined the association between small-scale greening and physiological stress responses, as well as self-reported psychological indicators (e.g., mood, anxiety, distress, perceived stress, recovery, attention, or cognitive function).
Studies did not examine the association between exposure to small-scale greening and physiological stress responses, as well as self-reported psychological indicators (e.g., mood, anxiety, distress, perceived stress, recovery, attention, or cognitive function).
Studies that use human participants Studies that do not use human participants Records were not written in English.
Records were not written in English.

Study Selection
Records from each database were downloaded and merged into Endnote X9. Duplicates were removed, and titles, abstracts, and full text were filtered. Study selection was conducted in three stages: the first stage involved screening the titles of the articles against the inclusion criteria; the second stage involved screening the abstracts; the third stage involved a full screening of the articles. In the first and second phases, screening was deferred to the next phase if there was doubt about the inclusion of any article. At this point, all team members were involved in discussions to resolve the issue. If there were duplicate studies, the authors randomly selected one from three databases.

Data Extraction
The first author (Jiaqi Gu) used a standardized form to extract the following information from each of the included studies: Article identifier (author name, year of publication, country); study identifier (setting, sample size, design); purpose of the study; and results. Results were extracted into a coding frame using Microsoft Excel, synthesized, and tabulated. A second researcher (Haixiao Liu) checked all data extraction tables, and disagreements were resolved through discussion until consensus was reached.

Quality Assessment
Each author independently assessed the risk of bias for each of the included studies. For the RCT, the risk of bias was assessed using the criteria outlined in the Cochrane Handbook for Systematic Reviews of Interventions [33]. The following areas were assessed against the Cochrane Risk of Bias Checklist: random sequence generation (selection bias), allocation concealment (selection bias), blinding of participants and personnel (performance bias), blinding of outcome assessment (detection bias), incomplete outcome data (attrition bias), selective reporting (reporting bias), and other bias performance. The risk of bias was assessed for each study-related entry. The risk of bias in each domain was divided into three categories: low risk, high risk, or unclear risk of bias.
The Risk of Bias in Non-Randomized Studies of Interventions (ROBINS-I) tool was used for non-randomized studies [34]: confounding, participant selection, classification of interventions, deviation from intended intervention, missing data, measurement of outcomes, reporting bias, and overall bias. The risk of bias in each area was divided into five categories: low risk of bias, moderate risk of bias, serious risk of bias, critical risk of bias, and no information.
Any differences are resolved through discussion.

Search Results
The key characteristics of each included study are detailed in Table 2. The initial search of the six databases produced a total of 2500 results. After removing 939 duplicate and substandard studies, a total of 1561 titles and abstracts were screened. A full-text search was also conducted for papers considered potentially relevant. The review flow chart is detailed in Figure 1.  The range of publication years for the included studies was from 1995 to 2021, with the majority of studies after 2010 (n = 13). These included studies were conducted in eight different countries. Ten of the nineteen studies were conducted in Asia, three in Europe, five in North America, and one in Australia. The countries with the highest frequency of inclusion were China and the USA, with ten studies (five studies each). Most studies were randomized controlled studies (n = 13) and the remainder were non-randomized controlled studies (n = 6) [5,12,16,[35][36][37][38][39][40][41][42][43][44][45][46][47][48][49][50].
The size of the sample surveyed varied considerably, with sample sizes ranging from 15 [38] to 240 [12]. Fifteen of these studies had a sample size of no less than 50. Participants were generally young, with 84% of the studies (n = 16) recruiting students (mean age < 30); two studies were working people and patients with a wide age range [37,48]; one study recruited senior citizens (range 61-97 years) [38]; and one study had identified a specific population: patients recovering from appendectomy [48]. Each of the nineteen studies had either male or female participants [45,46]. Fourteen articles had both male and female participants. Three studies did not describe the gender of the participants [39,44,47].

Risk of Bias
Thirteen randomized studies were evaluated using the standard [33] tools outlined in the Cochrane Handbook for the Systematic Evaluation of Intervention Systems, and six non-randomized studies were evaluated using ROBINS-I. (The overview of the risk of bias in the included studies is presented in Appendix A, Figures A1 and A2, and Table A1).
Of the thirteen randomized studies included, only one described their method of random assignment and was therefore designated as low risk of bias in this domain. The remaining studies did not describe their method of random sequence generation and were rated as unclear [35]. In terms of blinding, one study noted a double-blind trial [5]. Other RCTs in this review either did not provide information on participants and personnel or lacked blinding of participants and personnel and were assessed as unclear or at high risk of bias in this area. There were no participant dropouts in the five randomized studies, complete data were reported [12,38,45,46,48], and therefore the risk of bias was low. Four randomized studies with missing data were judged to be at high risk of bias [5,35,42,50]. The remaining randomized studies had incomplete information, so it was difficult to judge the completeness of the data, or the studies did not mention completeness issues and were therefore designated as unclear risk.
Of all included non-randomized studies, five were flagged as having an overall significant risk of bias. In terms of confounding factors, the four included in non-randomized studies were flagged as having a serious risk of bias, and many important confounding factors were identified (weather, food, physical activity, whether smoking and drinking alcoholic or caffeinated beverages before the experiment; study and homework stress; and plant wilt, etc.). For example, in the study by Ke-Tsung Han (2008) et al., participants may have been influenced by confounding factors such as exam stress and homework [33]; in the study by Virginia I. Lohr et al., (1996), participants' level of human-computer use and keyboarding skills were unclear, and thus both were noted as having a serious risk of bias [49]. One study was marked as moderately biased because participants were controlled for physical status as well as dietary intake [39].  Participants in the ward with indoor plants felt less self-reported stress than those in the ward with paintings. Indoor plants reduced stress by increasing the attractiveness of the room.  Sit in a chair for 5 min and walk around for 1 min The small green wall had a more dramatic improvement effect. Subjects in the small green wall condition had substantially lower stress levels compared to the non-green wall condition. The large green wall reduced STAI levels but increased mental stress, fatigue, and anxiety. Another study provided too little information on confounding factors and was marked as "no information" [47]. In terms of participant selection, all non-randomized studies were judged to have a low risk of selection bias due to the selection of participants in the included studies independent of the intervention or outcome. Judging by the character of the intervention, the risk of misclassification of intervention and control sites is low. Only one study was considered as a moderate "deviation from the intended intervention" because participants experienced unexpected plant death and the need to replace plants during plant care [37]. For the missing data domain aspect, four participants in one study were rated as having a serious risk of bias because they did not complete their study tasks resulting in missing data and were not addressed by appropriate analysis [44]. All other non-randomized studies had complete data or dealt with missing data and were judged to have low-risk bias. In terms of bias in the measurement of outcomes, all studies had data obtained through self-reporting, four studies were identified as having a severe risk of bias, and two studies were identified as having a moderate risk of bias [37,39]. All other non-randomized studies were at moderate risk of reporting bias due to reporting of complete data but no associated study protocol.
The most frequently used measure was blood pressure (n = 8). It was followed by the State Trait Anxiety Inventory (STAI, n = 6). Pulse rate (n = 5), electroencephalogram (EEG, n = 5), semantic differential (SD, n = 5), and heart rate variability (HRV, n = 4) were also used. Most studies were indoor greenery (n = 15), and others were small outdoor green spaces (n = 4). A small number of studies were greened for virtual scenes (n = 3) and scene photos (n = 2). In terms of environmental exposure, most studies (n = 15) did not have any green within the control environment. Two studies were controlled for environmental exposure to different areas of greenery [16,50], one study was controlled for different greenery behaviors and per capita area within a small green space [12], and one study was controlled for the presence or absence of odor from greenery [45].
The measurement methods and frequencies of the included studies are shown in Figure 2.

Effect on Blood Pressure
Five RCT studies [12,38,45,46,48] and three non-RCT studies [39,47,49] reported on the effects of small-scale green on blood pressure. An increase in blood pressure indicates an increase in stress [38]. Six studies showed a decrease in blood pressure and a decrease in stress in participants through interventions [12,[45][46][47][48][49]. One study showed no significant change in blood pressure for 5 min of ornamental plants in an indoor work environment [39]; another study on senior citizens in a small garden showed that one hour of outdoor recreation did not affect participants' blood pressure [38]. WeiLin et al., found no significant differences in blood pressure between the experimental groups after a green space behavior and per capita area intervention in six small green spaces and that the positive effect of the green environment on blood pressure was dynamic, with positive effects appearing in the first few minutes, then decreasing and disappearing in the long term [12].
Most studies were indoor greenery (n = 15), and others were small outdoor green spaces (n = 4). A small number of studies were greened for virtual scenes (n = 3) and scene photos (n = 2). In terms of environmental exposure, most studies (n = 15) did not have any green within the control environment. Two studies were controlled for environmental exposure to different areas of greenery [16,50], one study was controlled for different greenery behaviors and per capita area within a small green space [12], and one study was controlled for the presence or absence of odor from greenery [45].
The measurement methods and frequencies of the included studies are shown in Figure 2.

Effect on Blood Pressure
Five RCT studies [12,38,45,46,48] and three non-RCT studies [39,47,49] reported on the effects of small-scale green on blood pressure. An increase in blood pressure indicates an increase in stress [38]. Six studies showed a decrease in blood pressure and a decrease in stress in participants through interventions [12,[45][46][47][48][49]. One study showed no significant change in blood pressure for 5 min of ornamental plants in an indoor work environment [39]; another study on senior citizens in a small garden showed that one hour of outdoor recreation did not affect participants' blood pressure [38]. WeiLin et al., found no significant differences in blood pressure between the experimental groups after a green space behavior and per capita area intervention in six small green spaces and that the positive effect of the green environment on blood pressure was dynamic, with positive effects appearing in the first few minutes, then decreasing and disappearing in the long term [12].

Effect on Pulse Rate
The lower the pulse rate in the normal range, the less physiological tension and stress there will be [12]. Five studies included pulse as an assessment metric in intervention trials [12,37,40,45,49]. Studies have reported that both small outdoor green spaces and small indoor plants are conducive to lower pulse rates. WeiLin et al., claimed that walking in high PCA (per capita area) and sitting in low PCA had the most powerful effect on reducing stress.

Effect on HRV
Heart rate variability quantifies changes in heart rate associated with internal versus external environmental changes, particularly those due to autonomic nervous system activity. It is an indirect measure of mental stress and a key indicator of physiological stability [50]. Four RCT studies [16,40,46,50] reported the results of small-scale green for participant heart rate variability. In the study of the physiological and psychological effects of green walls on occupants, Seungkeun Yeom et al., concluded that heart rate variability is more sensitive to noise waves, and the measurement time is shorter and less reliable, thus discharging this indicator from the analysis. A report from 85 secondary school students showed that foliage plants resulted in a significant increase in parasympathetic activity (high-frequency component), suppression of sympathetic activity (low-frequency component), a significant decrease in pulse rate, and greater comfort and relaxation for the students upon seeing the plants [40]. Lee, MS (Lee, Min-sun) et al., claimed that the average activity in HRV (equivalent to sympathetic activity) decreased at the end of the vegetative task. Ji-Young Choi et al., reported that there was no significant difference in heart rate variability parameters among participants under conditions of 5%, 20%, 50%, and 80% green indices in indoor spaces; and that even very low green color provided physiological and psychological benefits [50].

Effect on Heart Rate
An RCT study reported a heart rate lowering effect of ward flowering foliage plants on patients undergoing abdominal surgery, with a lower heart rate implying a more positive physiological response [48]. A non-RCT study on senior citizens reported no effect of one hour of outdoor recreation on heart rate [38].

Effects on EEG
In this review, a total of five studies [16,39,45,47,50] included EEG as a basis for stress reduction assessment in their experiments. Generally, it is believed that alpha and beta waves show the closest relationship out of human emotions, with high alpha waves strongly associated with relaxation and high beta waves associated with attention and alertness [45]. Among the included studies, four of the RCT studies reported that exposure to smallscale green showed an increasing phenomenon of alpha and beta wave mean values on participants' EEG, but Seungkeun Yeom et al., showed a decrease in alpha and beta wave mean values under large green wall conditions, adding mental stress and fatigue as well as anxiety levels. A non-RCT study showed positive changes in EEG measurements after a small-scale greening intervention in indoor spaces, with no significant differences in EEG data between the different green indices, but there were gender differences, with men having significantly higher EEG power than women [50].

Effects on Anxiety
Two RCT studies [16,48] and four non-RCT studies [37,39,43,47] assessed the effects of small-scale greening on participants' status and anxiety levels. Five of the studies had significant reductions in STAI scores after the intervention [16,37,39,47,48], indicating that exposure to small-scale green was associated with significant improvements in anxiety levels and relief of psychological stress. A study was conducted on a specific population: post-appendectomy patients, and it showed that in-room plants were beneficial for patients' post-operative recovery. Seungkeun Yeom et al. found that the improvement effect of a large indoor green wall was much lower than that of a small green wall, leading to only a small decrease in state anxiety. In Ke-Tsung Han's study, the improvement in the anxiety state by six pots of leafy plants in the classroom was immediate, diminished with time, and statistically insignificant over a long period.

Effect on Emotion
Small amounts of greenery have a significant restorative effect on mood. Contact with greenery can be comforting [40,[45][46][47]50], natural and relaxing, and even small amounts of greenery can have a relaxing effect on people [50]. The Total Mood Disorder (TMD) and Profile of Mood States (POMS) measures found that participants gained more positive emotions and fewer negative emotions, with overall mood showing a more beneficial state [12,45]. One study showed that participants' Well-Being Measure (WBM) decreased over time [43].

Impact on Restorative Outcomes
Three studies reported a restorative effect of small-scale green interventions on participants [35,41,43], with Preyen Archery and Andrew Thatcher claiming a positive effect of the presence of houseplants on emotional recovery and no effect on cognitive recovery [41]. While another study concluded that there was no significant restorative effect of living plants and that visual richness in the environment may be a restorative factor [42].

Effects on Attention
Studies have reported that green roofs (virtual), outdoor gardens, and small plants in computer labs have an attention-improving effect [35,38,49]. However, in the Virginia I. Lohr et al. study, the specific exposure time was not stated, depending on the computer task completion time. WeiLin et al., claimed that six small green spaces had no significant improvement effect on college students' attention [12].

Impact on Self-Reporting
In one study, by having participants view photographs of a ward with or without plants, it was shown that indoor plants reduced participants' self-reported stress by increasing the attractiveness of the room [36]. In another study, it was shown that a 6-min plant viewing intervention increased participants' task engagement and reduced distress and worry [41].

Impact on Others
Thermal biofeedback is a method used to assess pressure by measuring the skin temperature of the fingers or toes. Cammie K. Coleman and Richard H. Mattson showed an increase in skin temperature during the first 5 to 8 min of a 20-min thermal biofeedback session, indicating stress reduction [44]. Another study aimed at patients recovering from surgery found that flowering plants in the ward had a relieving effect on patients' postoperative pain through a 101-point numerical rating scale (NRS-101) that assesses anxiety and fatigue [48].

Intervention Effectiveness
Of the 19 included studies, 84% (n = 16) reported varying degrees of reduction in participants' stress in the intervention condition. Ninety-five percent (n = 18) of the studies with small-scale greening did not report negative effects. Only one study reported a negative effect of small-scale greening: Seungkeun Yeom et al. 2021 reported that a large green wall increased participants' stress, fatigue, and anxiety levels [16]. Interventions and effectiveness results are shown in Table 3.

Effectiveness of Indoor Small-Scale Greening for Stress Reduction
Seven studies [39][40][41][42]45,49,50] evaluated the stress-reducing effects of short exposure (no more than 10 min) to small-scale indoor greenery. The presence of plants elicited significant changes in brain activity [39]. By measuring physiological stress such as blood pressure, pulse, and EEG, as well as psychological stress such as STAI and SD, the results of five of these studies showed a significant stress-reducing effect of small-scale greenery. Both indoor viewing greenery [39,40,45,50] and performing computer tasks in a small-scale greening environment [49] were beneficial for stress relief. Not only that, but participants also feel more relaxed and comfortable in a green environment [40] and even have increased productivity [49]. Two studies [41,42] did not report a stress-relieving effect of indoor greenery on participants' psychological stress. In addition, due to the lack of measurement of physiological stress, it was not possible to determine whether the interventions affected physiological stress.
The results of the three included studies [37,43,48] showed that prolonged (more than 1 day) exposure to small-scale indoor greenery had a mitigating effect on participants' psychological stress. Except for the study by Ke-Tsung Han (2008), which lacked a measure of physiological stress [43], the remaining two studies showed a significant reduction in physiological stress under small-scale greenery conditions [37,48]. Masahiro Toyoda et al. (2019) claimed that conscious gazing at nearby plants can reduce psychological and physiological stress in office workers [37]. Seong-Hyun Park and Richard H. Mattson (2008) showed a significant reduction in postoperative analgesic intake through an intervention in which patients exhibited more positive physiological responses (lower systolic blood pressure and heart rate) and more positive feelings [48].
A total of two studies [46,47] assessed the effects of plant activities on stress in subjects. Intervention results from both studies showed that a 15-min long plant transplant activity had a significant ameliorating effect on participants' psychological stress. Lee, MS (Lee, Min-sun) et al. (2015) concluded that positive interactions with indoor plants can reduce both physiological stress and psychological stress [46]. This is achieved by suppressing sympathetic nervous system activity and diastolic blood pressure as well as promoting a sense of comfort, soothing, and naturalness. The results of another study suggest that contact with plants can minimize mental stress [47].

Effectiveness of Small-Scale Outdoor Greenery on Stress Reduction
Of the 19 included studies, a total of two studies [12,38] assessed the effects of smallscale outdoor greenery on stress. With six outdoor small-scale simulated green spaces, WeiLin et al., (2019) reported that walking in high PCA (per capita area) and sitting in low PCA had the greatest reduction in stress [12]. Another study showed that senior citizens in a home for the elderly relaxing in an outdoor garden for one hour had only a restorative effect on attention. There was no effect on physical stress, while data on the measurement of psychological stress were lacking. Thus the small-scale greening in this study had no stress-reducing effect [38].

Effectiveness of Non-Living Plant Environments on Stress Reduction
A total of two studies [16,35] had small-scale greening as a virtual scenario. Both studies reported that small-scale greenery had a reducing effect on participants' stress to varying degrees. One of the studies, Seungkeun Yeom et al., (2021), showed a substantial reduction in stress levels in subjects with small indoor green wall conditions. However, when the indoor green wall was larger, it increased the mental stress, fatigue, and anxiety levels of the subjects [16]. Subjects in another study had elevated subcortical arousal and cortical attentional control after micro-breaks from viewing a simulated view of a green roof. The restoration of attention brought benefits, including improved performance and mood, as well as reduced stress [35].
A study by Bin Jiang et al. found that viewing small-scale greening videos reduced stress (2014). The intervention condition's measures of physical and psychological stress revealed a significant effect of community street tree cover density on stress reduction. However, the effects of different tree cover densities on stress reduction differed significantly by gender. Males' stress recovery dose curves showed an inverted U-shape, with faster stress recovery at 1.7-24% tree cover density, no significant change at 24-34%, and slower recovery above 34%, with moderate tree density causing greater stress reduction. Furthermore, there was no significant difference in pressure recovery for different tree densities among females [5].
Out of the 19 included studies, two showed an attenuating effect of plant pictures on stress [36,44]. According to one study, when people saw pictures of hospitals with plants, their levels of stress were lower than when they saw pictures of hospitals without plants. Indoor plants eased the tension by making the space more appealing [36]. Another study used plant color photography or live plants for the intervention. According to the study's findings, 38% of participants responded favorably to both actual plants and photos.

Discussion
In this systematic review, 19 studies from eight countries were synthesized. These studies investigated the impact of small-scale greening (photos, videos, or virtual environments) on stress relief.
Most studies use a holistic approach, collecting subjective psychological and objective physiological data to assess stress status. Pulse, blood pressure, HRV, EEG, STAI, and SD/SDM were frequently used to assess stress in 19 studies. Psychological stress is measured by subjective rating scales such as the STAI; physiological stress is usually measured by this analysis due to the validity, reliability, and ease of collection of pulse, blood pressure, HRV, and EEG data.
Our findings suggest that all of the included studies reported that varying degrees of small-scale greening are beneficial for health. Most of the studies (n = 15) demonstrated the mitigating effect of small-scale greening on physical or psychological stress as measured by various indicators. At the same time, we found that the intervention produced beneficial results such as restored concentration, pain relief, and feeling more comfortable and natural. Of the literature included in this review, a small number of studies reported interactions with plants, such as plant transplantation [46,47] and care [37], nine studies for directed attention to greenery, and seven studies for direct exposure to environments with greenery [12,38,[43][44][45]48,49]. The results of the study found that all three intervention types were helpful for stress relief, and due to the different number of included literature and insufficient data, it was not possible to determine which type was most helpful for stress reduction. However, it is unclear whether uncontrollable factors such as plant wilting during plant care can adversely affect stress reduction and needs to be studied in depth through other literature. An included study on primroses found that scented primroses were more effective in reducing stress than unscented ones. The authors concluded that the effect of plants on people is not only in the visual perception but that plants with aromatic properties can induce physiological and psychological relaxation through olfactory channels [45].
The immediate effect of plant stress reduction was significant. Studies have found significant stress reduction effects for short periods of exposure to small-scale green environments, consistent with the positive physiological responses to plants described in previous studies over 3 to 6 min [37,44]. Some studies concluded that prolonged exposure to stress, although improved, has a statistically insignificant effect [43].
The stress-reducing effect of greening facilities is not proportional to their size. When the indoor green wall is too large, it not only loses its effectiveness in reducing stress but also has a negative psychological impact. A small outdoor green space study included in this systematic evaluation showed the strongest effect in reducing stress when walking in high PCA (per capita area) and sitting in low PCA, but there is a lack of relevant data on whether other experimental settings are effective in stress reduction. Stress reduction theory suggests that the presence of a natural environment (restorative environment) brings about a psychological, evolutionary response related to safety and survival and therefore generates positive emotions. However, larger areas of green space are not our only goal, and health benefits can only be maximized when the area of green space per capita is matched with appropriate green space behavior. In the planning and design of urban green spaces, designers should provide more appropriate green space structures and layouts and advocate addressing the uneven distribution of urban green spaces rather than blindly pursuing larger green spaces [12,16,51].
A lower greening index and very small-scale greening may also have a positive impact on stress reduction. Ji-Young Choi et al. found no significant differences in physiological parameters among the four (5%, 20%, 50%, and 80%) green index levels studied, although participants subjectively preferred a 50% indoor green index. Another study found that stress recovery was faster when tree cover density was between 1.7% and 24% and that tree density above 24% was associated with slower recovery times, but the results were limited to males. For women, there was no relationship between different densities of tree cover and stress recovery. Studies have reported finding that even small, single indoor plants can people exert stress-reducing effects during conscious or unconscious gaze [37,39,50].
Outdoor green space is not the only thing that works for stress relief. Indoor greenery is another important way for people to get in touch with nature. For people who have less time to get in touch with outdoor green spaces and parks, contact with indoor plants can effectively reduce stress and improve work efficiency. Green facilities in hospital wards can bring more positive physical and psychological feelings, reduce pain, and relieve psychological stress for patients recovering from some diseases after surgery [36,37,39,42,43,[46][47][48][49].
The heterogeneity of the studies included in this review is high. Due to the nature of the studies, participants were likely aware of the hypotheses being tested. Most studies, therefore, showed a high risk of bias in terms of blinding, and there is no way to avoid some subjectivity in the results. In addition, some studies [32,38,43] were of long duration, and many uncontrollable factors may have influenced the experimental results. In order to strengthen the clinical evidence base, it is advised that future studies in this field use more stringent and standardized procedures to reduce bias and increase quality. This will help convince policymakers and health professionals that even small-scale greening can have a beneficial impact on stress mitigation.

Limitations
Our review has several limitations. The keywords used in our search strategy for identifying eligible studies may have been narrow due to the lack of a standard definition of small-scale greening and inadvertently excluded greening that may have met the inclusion criteria size. Most of the exposure environments in the included literature were indoor environments, and data from studies of small-scale outdoor greening were lacking, so the findings lacked greater rigor and accuracy. There is a wide variety and poor uniformity of methods regarding pressure testing in the included literature. Therefore, a relevant meta-analysis could not be performed. Many studies suffer from small sample sizes and single study populations, and the results are not generalizable. Most of the included studies were single experiments with an intervention length of 15 min or less, and the findings could only demonstrate the effect of greening in a limited time task, and it was not possible to determine whether the same effect on stress existed with long-term exposure. A small number of exposures to non-living plant environments (photographs and virtual scenes) are also included in this review, and although supported by relevant research theories, their ability to replace real environments remains controversial. Soil and water from indoor plants are potential sources of bacteria, and we cannot rule out the risk of harmful health effects from indoor landscaping and the impact of airborne particles produced. There are many uncontrollable factors during long experiments, such as academic stress, diet, sleep, etc., resulting in a lack of rigor in the results. The included study populations were mostly young students, which limits the generalizability of our findings. In addition, the research literature in this review is all English-language studies, which should be screened more comprehensively.

Conclusions
In this systematic review, 19 studies from 8 countries were synthesized. These studies investigated the impact of small-scale greening (photos, videos, or virtual environments) on stress relief.
Although more research exists on the benefits of natural environments for stress reduction, there is a lack of systematic research on the role of small-scale greenery in reducing stress. This systematic evaluation helps to fill this knowledge gap and suggests that small-scale greenery, including green walls and potted plants, may have beneficial effects on stress reduction. Understanding the physiological and psychological benefits of small-scale greening could help better inform urban design planning in the context of dense urban trends that provide more opportunities for urban residents to engage with nature. However, the research is highly biased and of low quality, and more rigorous studies are needed to improve our understanding of the relationship between greening and stress reduction.