Association between Active Use of Urban Green Spaces and Well-Being in Adults Aged 18–65 Years: A Systematic Review

Abstract

Background:

Engaging in physical exercise in urban green spaces is increasingly recognized as a crucial factor for well-being, offering a range of benefits for physical, mental, and social health in growing urban environments.

Objective:

This systematic review aimed to analyze the scientific literature exploring the correlation between participating in physical exercise and sports activities in urban green spaces and the well-being of participants, encompassing physical, psychological, mental, and social aspects.

Methods:

This systematic review adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. The search strategy included articles published in English, Spanish, or Portuguese within the databases Medline (PubMed), Scopus, and Web of Science (WoS) up to 15 June 2023. Studies were selected based on predefined inclusion and exclusion criteria, which included descriptive and observational epidemiological study designs and a population of adults 18–65 years of age. The Effective Public Health Practice Project (EPHPP) tool was used to assess the risk of bias in the included studies.

Results:

A total of 6,634 studies were identified, of which 31 met all the inclusion criteria. The review identified evidence suggesting that physical exercise in urban green spaces can have a positive impact on health, including improvements in diastolic blood pressure and cardiovascular health, such as heart rate, heart rate variability, cardiac and vascular function, as well as mood and self-esteem. In addition, evidence supports an association with decreased stress and a lower risk for mental health conditions, such as anxiety and depression, in addition to a strengthening of social cohesion and social interaction.

Discussion:

This systematic review suggests that physical exercise in urban green spaces is associated with positive health outcomes. The findings support a link between such activities and improvements in physical, psychological, mental, and social well-being. Despite variations in defining urban green spaces, the results highlight the importance of promoting physical activity programs in these environments to enhance the overall well-being of urban populations. https://doi.org/10.1289/JHP1040

Introduction

Urbanization is a globally expanding phenomenon, with more than half of the world’s population now residing in urban areas.¹ As our cities grow, the availability of green spaces diminishes, making the remaining green spaces increasingly essential for human well-being.^2,3 The definition of “urban green spaces” remains a topic of active debate across various disciplines, with no universally accepted definition.⁴ This ambiguity reflects the diversity of functions and characteristics that these spaces can have, making a singular interpretation challenging. Despite this debate, in the present study we adopt an operational definition that describes urban green spaces as publicly owned areas, accessible to the public, and characterized by a high degree of vegetative cover, ranging from parks and forests to natural areas,⁵ streams, riverbanks, and waterfront zones.⁶ They not only provide a respite from the hustle and bustle of urban life but can also foster physical exercise, recreation, and social interaction.⁷

Modern urban living is often associated with high levels of stress, sedentary behavior, and mental and physical health issues.⁸ The lack of opportunities for physical activity in urban environments has contributed to this concern, and it is in this context that urban green spaces play a crucial role.⁹ In this scenario, the relationship between engaging in physical and sports activities in urban green spaces and individual well-being has emerged as a critically important topic in contemporary research.¹⁰ A particularly relevant area is the concept of green exercise, defined as physical activity performed in green and natural environments.^11–13

Although intuition suggests that physical activity in urban green spaces should be beneficial for health and well-being, precise and unified scientific evidence supporting this notion is still partial and sometimes contradictory. Some reviews have reported significant positive associations with the psychological and mental health^12,14,15 outcomes of individuals engaging in outdoor activities in urban green environments, even comparing these positive effects to indoor or urban environments,^11,16,17 while others have presented more modest or opposing findings.^18,19

With increasing urbanization and growing concerns about health and well-being in urban environments, this review aimed to play a crucial role by consolidating and synthesizing current research in this field. Although existing systematic reviews have examined the effects of physical activity in green spaces on physical¹⁸ or psychological^11,18,19 well-being as primary outcomes, a comprehensive review that considers all aspects of well-being (physical, psychological, mental, and social) is currently lacking. This review aimed to analyze the scientific literature exploring the relationship between engaging in physical and sports activities in urban green spaces and the full spectrum of well-being for participants.

Methods

This systematic review was conducted following the guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA)²⁰ and was registered in the prospective international register of systematic reviews (PROSPERO-CRD42023436963). Article searches were performed in the following databases: Medline (PubMed), Scopus, and Web of Science (WoS), up to 15 June 2023, with no restrictions on publication date or status.

Eligibility Criteria

The eligibility criteria selected for the articles were based on inclusion and exclusion criteria. Studies were included if they

1. Investigated the relationship between engaging in some form of exercise (e.g., walking, running, cycling) in urban and peri-urban green spaces and its impact on one or more dimensions of well-being: physical, psychological, mental, and social.

2. Used descriptive and observational study designs, including cross-sectional, experimental, or longitudinal designs.

3. Included a population of adults 18–65 years of age.

4. Had no restrictions regarding the country of origin.

5. Were published in English, Spanish, or Portuguese.

Exclusion criteria were applied to studies that

1. Did not specify the type of physical or sports activity in urban and peri-urban green spaces, excluding contemplative or exposure activities.

2. Were based on materials such as literature reviews, systematic reviews, unpublished data, gray literature, unpublished theses, duplicated studies, books, conference communications, editorials, comments, case reports, or studies with confusing or likely erroneous data.

3. Did not provide information about the study population or included individuals $<18\text{\hspace{0.17em}or}>65$ years of age.

4. Were not available in full.

5. Were not published in English, Spanish, or Portuguese.

Search Strategy

The searches were conducted in the following databases: PubMed, Scopus, and WoS. The search strategy involved a combination of terms related to urban green spaces, exercise, and well-being. For the identification of terms related to urban green spaces, the relevant terms used in the systematic review conducted by Rojas-Rueda²¹ were selected. Terms related to exercise were selected based on the systematic review conducted by Thompson Coon.¹⁹ Finally, for terms related to well-being, synonyms in English and terms covering physical, psychological, mental, and social aspects of well-being were included. This strategy listed below was designed with the aim of ensuring thoroughness and precision in the retrieval of relevant literature for the systematic review in question:

(“recreational park” OR “urban green space” OR “green space” OR “open space” OR “public open space” OR “park” OR “city park” OR “public park” OR “public garden” OR “urban park” OR “pocket park” OR “municipal space” OR “municipal park” OR “municipal land” OR “public land” OR “built environment” OR “greenway” OR “urban greenway” OR “green infrastructure”) AND (“exercise” OR “nature-based exercise” OR “recreation” OR “physical activity” OR “physical exercise” OR “Leisure activities” OR “leisure activity” OR “Activities” OR “walking” OR “green exercise” OR “outdoor training” OR “outdoor walks” OR “brief walks” OR “running” OR “Bicycling” OR “cycling”) AND (“wellbeing” OR “well being” OR “well-being” OR “psychological well-being” OR “physical fitness” OR “physical health” OR “social cohesion” OR “social well-being” OR “body mass index” OR “psychological stress” OR “psychological stressor” OR “mental health” OR “mental wellbeing”).

The search strategy was adjusted when searching across different databases (Table S1).

Study Selection

The study selection was conducted by retrieving information from databases and then adding it to Mendeley software for the removal of duplicate articles.²² Subsequently, all articles were incorporated into the Rayyan tool (http://rayyan.qcri.org).²³ Two authors, independently (J.L.H. and J.F.G.V.), manually reviewed the titles and abstracts of all articles using the Rayyan software²⁴ to select those that met the inclusion criteria. In case of disagreement, a third researcher (L.F.J.G.C.) joined the discussion to resolve any discrepancies. Finally, the selected articles based on titles and abstracts underwent a review of the full text. Articles that met the study objectives were included in the analysis, whereas the others were excluded.

Data Collection Process

The key data were extracted from each selected article using a custom form. This form included details, such as author and year, country, study design, sample size, age, population, description of the urban green space, type of physical activity, well-being measures assessed, and the main results. This process ensured a systematic collection of information for a comprehensive assessment of the relationship between physical activity in urban green spaces and human well-being.

Assessing the Risk of Bias

The Effective Public Health Practice Project (EPHPP) tool was employed to assess the quality of the included studies.²⁵ This tool is applicable to various study types, including observational studies, cross-sectional studies, pre/post studies, and randomized controlled trials.²⁶ It includes six categories—selection bias, study design, confounding factors, blinding, data collection methods, and withdrawals/dropouts—with each category rated as strong, moderate, or weak according to specific criteria. For instance, a “strong” rating in selection bias indicates representative sampling methods, whereas a “weak” rating suggests possible bias due to nonrepresentative samples. The overall rating is then based on the category scores: strong (no weak ratings), moderate (1 weak rating), and weak ($\ge 2$ weak ratings). Two researchers (J.L.H. and J.F.G.V.) independently conducted and documented risk of bias assessments for each study across all domains. In cases where consensus between these two researchers could not be reached on a domain of risk of bias, another researcher (A.R.P.Z.) reviewed the results.

Results

Through the electronic database search, 7,916 documents were identified (Figure 1). After removing duplicates ($n=\mathrm{1,282}$), 6,634 potentially relevant documents remained. Title and abstract screening excluded 6,478 documents that did not meet the specific inclusion criteria for the study (e.g., not focused on physical activity in urban green spaces, not measuring well-being outcomes). Of the initial 156 articles selected for full review, 125 were subsequently excluded. Reasons for exclusion included no well-being measure identified ($n=19$), not meeting the condition of physical activity and/or sports ($n=60$), not meeting the condition of urban green space ($n=6$), not meeting the criteria for age ($n=1$), and other reasons that became apparent during full review that did not align with the study’s specific objectives ($n=39$). The total number of included articles meeting all criteria was 31.

Figure 1.

Flow diagram showing how documents were selected for the study.

Study Characteristics

The 31 studies spanned four continents, were conducted in 13 countries, and were published from April 2003 to September 2022. In Asia, 3 studies were from Malaysia,^27–29 3 from Taiwan,^30–32 4 from Japan,^33–36 1 from China,³⁷ and 1 from Bangladesh.³⁸ In Europe, 6 studies were from the UK,^39–44 3 from Finland,^45–47 2 from Sweden,^48,49 1 from Switzerland,⁵⁰ and 1 from Ukraine.⁵¹ In North America, 3 studies were from the United States^52–54 and 1 from Canada.⁵⁵ In Oceania, 2 studies were from Australia.^56,57

Risk of Bias

We used the quality assessment described for the EPHPP approach to conducting systematic reviews to evaluate the risk of bias for all studies included in this review.²⁵ Except for observational studies, most studies had small sample sizes. The lowest scores were obtained in the categories of selection bias ($n=16$), blinding ($n=9$), and study design ($n=8$). The categories of confounding factors ($n=17$), blinding ($n=8$), and withdrawals and dropouts ($n=5$) were not applicable for select studies owing to the observational or cross-sectional trial design with a single group. Finally, the overall ratings revealed that 6 studies were rated as strong, 14 as moderate, and 13 as weak (Table 1).

Table 1.

Risk of bias assessment: EPHPP tool.²⁵

Study/year	A	B	C	D	E	F	Results
Bailey et al. 201852	M	M	NA	W	M	S	Moderate
Barton et al. 200939	W	M	W	W	S	NA	Weak
Bodin and Hartig 200348	W	M	NA	W	S	S	Weak
Brown et al. 201440	M	S	S	M	S	M	Strong
Flowers et al. 202256	W	M	NA	M	S	S	Moderate
Geniole et al. 201655	S	S	S	W	S	S	Moderate
Gladwell et al. 201641	W	S	S	S	S	S	Moderate
Hakman et al. 202151	W	M	NA	W	S	S	Weak
Han 201730	W	S	S	M	S	S	Moderate
Han 202131	S	S	S	M	S	S	Strong
Holt et al. 201953	S	W	NA	NA	S	S	Moderate
Johansson et al. 201149	M	M	S	W	S	S	Moderate
Kajosaari and Pasanen 202145	W	W	NA	NA	M	W	Weak
Kinnafick and Thøgersen-Ntoumani 201442	S	M	S	M	S	S	Strong
Kinnafick and Thøgersen-Ntoumani 201442	S	M	S	M	S	S	Strong
Lanki et al. 201746	S	M	S	W	S	S	Moderate
Lee et al. 202154	W	W	NA	NA	M	S	Weak
Mao et al. 202237	M	M	S	NA	W	S	Moderate
Marselle et al. 201343	W	W	NA	NA	S	NA	Weak
Martens et al. 201150	W	S	S	M	S	S	Moderate
Mokhtar et al. 201827	W	M	NA	W	S	S	Weak
Pasanen et al. 201847	S	S	S	M	S	S	Strong
Pasanen et al. 201847	M	S	S	S	S	W	Moderate
Paul et al. 202038	M	W	NA	NA	W	NA	Weak
Pretty et al. 200744	S	W	NA	NA	S	NA	Moderate
Rajoo et al. 202128	M	S	S	M	S	M	Strong
Raman et al. 202129	S	M	NA	W	S	S	Moderate
Shanahan et al. 201657	S	W	NA	NA	S	NA	Moderate
Song et al. 201333	W	S	NA	M	W	S	Weak
Song et al. 201434	W	S	NA	M	W	S	Weak
Song et al. 201535	W	S	NA	M	W	S	Weak
Song et al. 202236	W	S	NA	M	W	S	Weak
Tsao et al. 202232	W	M	W	M	M	S	Weak

Note: Some studies could not be rated in certain categories owing to study design limitations. For example, the categories of confounding factors (C) and blinding (D) were not applicable in certain observational studies or studies without control groups, where confounding factors could not be controlled or blinding was not feasible. A, selection bias; B, study design; C, confounding factors; D, blinding; E, data collection method; EPHPP, Effective Public Health Practice Project; F, withdrawals and dropouts; M, moderate; NA, not applicable; S, strong; W, weak.

Population

The main characteristics of the included studies are summarized in Table 2. The sample for this study consisted of 31 different investigations, with a total of 7,863 participants. Two of the articles presented the results of 2 different studies.^30,42,47 The reviewed studies revealed that 12 of them focused on university student populations,^{27,30,31,33–37,49,52,53,56} members of the university community,⁴² or a mix of students with full-time working individuals,⁵⁰ and 10 studies focused on the adult population,^{32,38,39,41,43,44,47,51,55,57} although some studies focused on a specific group. For example, Rajoo et al.²⁸ focused on a young adult population 18–40 years of age in Malaysia; Lee et al.⁵⁴ and Bodin and Hartig⁴⁸ targeted recreational runners; and Brown et al.⁴⁰ focused on middle-aged office workers with an average age of 42 y in the UK. Raman et al.²⁹ focused on service-oriented workers, 25–60 years of age, in Malaysia. Some studies focused on only male populations, for example, Geniole et al.⁵⁵ with adults, or Song et al.^33–36 and Mokthar et al.,²⁷ who targeted university students in Japan and Malaysia, respectively, whereas other studies focused on only the female gender, for example, one with an adult population as the sample in Finland.⁴⁶

Table 2.

Physical exercise in urban green spaces (UGS) and well-being.

Author, year, and country	Study design, sample size, gender, age, population	Description of the type of urban green space	Exercise	Evaluative measures of well-being	Main findings
Physical well-being
Brown et al. 2014; UK40	Cross-sectional; $\mathrm{P}=73$; M: ND, F: ND; $\text{mean age}=$ 42 y; office workers.	Natural environment consisting of trees, well-maintained grass areas, public footpaths, and rural roads. Built trail mainly composed of paved pathways adjacent to roads, residential areas, and industrial zones.	Walking.	SF-8 health questionnaire: self-rated general health. HR, natural logarithm of high frequency (lnHF) to explore changes in HRV. SBP and DBP. Framingham Cardiovascular Disease (CVD) risk score (%). Chester Step test. Body mass index (BMI).	Walking in the natural environment improved DBP ($p<0.01$) compared with the built environment and the control group. No significant differences were found in any of the other health measures assessed (self-rated health, physical health, BMI, predicted fitness, CVD risk, HR, lnHF, HRV, and SBP).
Geniole et al. 2016; Canada55	Cross-sectional; $\mathrm{P}=31$; M: 31, F: 0; $\text{mean age}=$ $24.61\pm 3.88$ y; adult males.	Naturalized landfill (designed to resemble a natural and green park) and urban area (commercial and business zone).	Walking.	Salivary cortisol and testosterone.	There was no significant difference between walking in the naturalized landfill and the urban environment in cortisol ($p=0.62$) and testosterone ($p>0.17$).
Gladwell et al. 2016; UK41	Cross-sectional; $\mathrm{P}=13$; M: 7, F: 6; $\text{mean age}=$ $39.4\pm 13.9$ y; adult population.	Green space on the university campus (grasslands, wooded areas, and a small lake) and urban environment (student residences, roads, shops, and other campus buildings).	Walking.	HRV: mean RR interval data (with SDNN) for sympathetic activity and rMSSD for parasympathetic activity and Poincaré plot SD1.	The indicators of HRV [SDNN (95% CI: 0.0, 13.1: $p=0.039$), rMSSD (95% CI: $-0.1$, 17.3: $p=0.041$), and SD1 (95% CI: 0.0, 12.3: $p=0.041$)] during sleep after walking in the natural environment were significantly higher than in the built environment. There was no significant difference in RR or sleep duration between environments.
Hakman et al. 2021; Ukraine51	Cross-sectional; $\mathrm{P}=101$; M: 50, F: 51; range: 60–65 y; adult population.	Urban park.	Physical exercise sessions (warm-up: stretches; main part: aerobic exercises (walking + strength exercises); cool-down: breathing exercises + stretches).	Framingham Physical Activity Index PWC150 test. SF-36: physical health, self-rated general health.	Physical activity index increased in both men and women. Physical activity was $81.3\pm 0.6$ points in men and $85.8\pm 0.6$ points in women, general health was $76.4\pm 0.3$ points in men and $75.4\pm 0.3$ points in women. The results indicate an improvement in work physical capacity (PWC150 test: $p=0.05$).
Holt et al. 2019; USA53	Cross-sectional; $\mathrm{P}=207$; H: 31.4%, M: 69.6%; age: ND; university students.	University campus featuring a combination of open green spaces and wooded areas.	Hiking/walking, running, cycling.	Self-rated general health: “Would you say that, in general, your health is…?”	Active use of green spaces improves general health ($p<0.01$).
Lanki et al. 2017; Finland46	Cross-sectional; $\mathrm{P}=36$; H: 0, M: 36; $\text{mean age}=$ $46\pm 8.7$ y; range: 30–60 y; female adult population.	Urban forest, urban park, and city center.	Walking.	Blood pressure (SBP, DBP). HR. Electrocardiogram: HRV (SDNN, RMSSD, HF).	DBP increased in natural environments (forest: $p=0.008$, park: $p=0.018$, city: $p=0.387$); SBP increased, but there was no significant difference (forest: $p=0.066$, park: $p=0.112$, city: $p=0.119$). HR decreased in green environments (forest: $p=0.041$, park: $p<0.001$) but not in urban (city: $p=0.126$). For HRV, SDNN increased in all three environments (forest: $p<0.001$, park: $p<0.001$, city: $p<0.001$), and RMSSD and HF decreased in the city ($p<0.001$) but were not affected by green environments.
Mao et al. 2022; China37	Longitudinal; $\mathrm{P}=\mathrm{2,852}$; M: ND, F: ND; $\text{mean age}=$ 18.35 y; university students.	University campus with green coverage and trees.	Running.	National Student Physical Health Standard of China (NSPHS).	Significant differences were seen in changes of the physical health in NSPHS tests among students ($p<0.05$), reaching a peak in the group with a total mileage of 120–140 km ($\mathrm{\beta }=0.101$, $p<0.01$).
Mokhtar et al. 2018; Malaysia27	Cross-sectional; $\mathrm{P}=20$; M: 20, F: 0; $\text{mean age}=$ 21.1 y; male university students.	Urban park and urban environment.	Walking.	Salivary cortisol, blood pressure, HR (physiological measures to gauge stress).	Cortisol levels showed a significant difference compared with the UGS setting, which is lower than in the city (UGS: $0.89\pm 0.55$; city: $2.33\pm 1.04$; $p<0.05$). DBP showed a decrease in UGS (pre: $77.9\pm 5.53$; post: $70.5\pm 10.05$; $p<0.05$) and lower values in UGS than in the city (city: $76.6\pm 10.69$; $p<0.05$). There was no significant difference in SBP in both environments. Significantly lower HR values in UGS compared with the city (UGS: $66.8\pm 10.71$; city: $72.4\pm 13.71$; $p<0.05$).
Paul et al. 2020; Bangladesh38	Cross-sectional; $\mathrm{P}=181$; M: ND, F: ND; $\text{mean age}=$ 47 y; adult population.	Open public spaces.	Walking/jogging, cycling, exercising, socializing with friends, relaxing, accompanying children, playing (soccer, skating).	Structured questionnaire that included a variety of well-being considerations: physical health.	Perception regarding improvement in physical health: Help in reducing or maintaining body weight: ${\mathrm{\chi }}^2=27.65$; $p<0.001$. Help in reducing or maintaining hypertension: ${\mathrm{\chi }}^2=155.05$; $p<0.001$. Help in reducing cholesterol level: ${\mathrm{\chi }}^2=161.81$; $p<0.001$. Help in reducing blood sugar level: ${\mathrm{\chi }}^2=118.35$; $p<0.001$. Help in reducing liver fat: ${\mathrm{\chi }}^2=290.54$; $p<0.001$.
Shanahan et al. 2016; Australia57	Cross-sectional; $\mathrm{P}=\mathrm{1,538}$; M: ND, F: ND; range: 18–65 y; adult population.	Public green spaces.	Walking.	Information on treatment for high blood pressure to measure physical health.	Higher frequency of physical activity was not correlated with lower hypertension, with a pseudo $R^2$ of 0.06 ($p<0.04$).
Song et al. 2013; Japan33	Cross-sectional; $\mathrm{P}=13$; M: 13, F: 0; $\text{mean age}=$ of $22.5\pm 3.1$ y; male university students.	Urban park and city area.	Walking.	HR to measure relaxation, HRV to measure relaxation.	HR values were significantly lower (4.4%) after walking in the urban park in winter ($98.4\pm 0.9\text{\hspace{0.17em}bpm}$) than after walking in the city area ($102.9\pm 1.1\text{\hspace{0.17em}bpm}$; $p<0.05$). A significant difference in HRV was found in the natural logarithm of HF (ln(HF)), after walking in the urban park in winter ($4.61\pm 1.25$ ${\text{ms}}^2$) showed a 21.6% higher value than after walking in the city area ($3.79\pm 1.16$ ${\text{ms}}^2$; $p<0.05$).
Song et al. 2014; Japan34	Cross-sectional; $\mathrm{P}=17$; M: 17, F: 0; $\text{mean age}=$ $21\pm 1.7$ y; male university students.	Urban park and city area.	Walking.	HR to measure relaxation, HRV to measure relaxation.	The mean HR was significantly lower (4.0%) during walking in the urban park in spring than during walking in the city area (UP: $86.7\pm 2.9\text{\hspace{0.17em}bpm}$, CA: $90.3\pm 2.6\text{\hspace{0.17em}bpm}$; $p<0.05$). For HRV, ln(HF) was 17.1% higher during walking in the urban park in spring than during walking in the city area (UP: $4.1\pm 0.2$ ${\text{ms}}^2$, CA: $3.5\pm 0.2$ ${\text{ms}}^2$; $p<0.01$).
Song et al. 2015; Japan35	Cross-sectional; $\mathrm{P}=23$; M: 23, F: 0; $\text{mean age}=$ $22.3\pm 1.2$ y; male university students.	Urban park and city area.	Walking.	HR to measure relaxation, HRV to measure relaxation.	The average HR during walking was significantly lower during the walk in the urban park in fall than during the walk in the city (UP: $86.4\pm 1.6\text{\hspace{0.17em}bpm}$; CA: $89.7\pm 1.6\text{\hspace{0.17em}bpm}$; $p<0.01$). For HRV, the mean ln(HF) during the entire walking period was significantly higher in the urban park in fall than in the city area (UP: $4.1\pm 0.2$ ${\text{lnms}}^2$; CA: $3.6\pm 0.2$ ${\text{lnms}}^2$; $p<0.01$).
Song et al. 2022; Japan36	Cross-sectional; $\mathrm{P}=51$; M: 51, F: 0; $\text{mean age}=$ $22.2\pm 2.0$ y, male university students.	Urban park and city area.	Walking.	HR to measure relaxation, HRV to measure relaxation.	In summer, HR was significantly lower when walking in an urban park than when walking on a city street ($92.3\pm 2.3$ vs. $96.1\pm 2.7\text{\hspace{0.17em}bpm}$, $p<0.01$). For HRV, differences in ln(HF) when walking in an urban park were significantly greater in spring, fall, and winter compared with in summer (spring: $0.60\pm 0.18$ vs. summer: $-0.14\pm 0.13$, $p<0.05$; fall: $0.51\pm 0.08$ vs. summer, $p<0.05$; and winter: $0.94\pm 0.46$ vs. summer, $p<0.01$).
Tsao et al. 2022; Taiwan32	Cross-sectional; $\mathrm{P}=25$; M: 18, F: 7; mean age of forest $\text{participants}=54.21$ y, mean age of urban forest $\text{participants}=42.36$ y; adult population.	Forest park and urban park.	Walking.	SBP and DBP. Central SBP and DBP. HR. Maximum LV pressure slope (dp/dt max). Cardiac output. Systemic vascular compliance and brachial artery compliance.	For blood pressure, there was no significant difference between the woodland park and the urban park in blood pressure but there was in pulse pressure between woodland and urban park ($p<0.0001$). Significant differences in HR values between woodland park and urban park ($p<0.0001$). Significant difference between woodland park and urban park for LV dp/dt max ($p<0.0001$) but not for cardiac output. Vascular function: significant difference in systemic vascular compliance ($p<0.0001$) and in brachial artery distensibility ($p<0.0001$) between woodland parks and urban parks.
Psychological well-being
Bailey et al. 2018; USA52	Cross-sectional, $\mathrm{P}=10$; M: 5, F: 5; $\text{mean age}=$ 20 y, university students.	Nature trail on campus (outdoors) and the aquatic recreation center (indoors).	Walking.	Stroop test. BDST. EEG measures to assess cortical activity corresponding to psychological states (attention, motivation, meditation, relaxation, anxiety).	Walking in a natural environment was associated with greater observed mental restoration, significant improvement in cognitive performance, as measured through the Stroop test ($p=0.042$) but not by BDST ($p=1.000$). EEG measures showed prolonged maintenance of relaxed mental states ($p=0.033$) and meditative states ($p=0.58$) after outdoor walking.
Barton et al. 2009; UK39	Cross-sectional; $\mathrm{P}=132$; M: 60, F: 72; range: 19–65 y; adult population.	Natural and heritage green spaces: Dunwich Heath, Suffolk; Flatford Mill, Suffolk; Hatfield Forest, Essex; Wicken Fen, Cambridgeshire.	Walking.	RSE. POMS. TMD.	Walking in green spaces had a positive impact on self-esteem ($p=0.0325$) and improved mood (POMS) ($p<0.0005$) and overall mood state (TMD) ($p<0.0005$).
Bodin and Hartig 2003; Sweden48	Cross-sectional; $\mathrm{P}=12$; M: 6, F: 6; adult population of men and women 26–46 years of age with $>3$ y of running experience.	Park trail (natural environment) and urban route (on sidewalks and streets, near a commercial area and heavy traffic).	Running.	PRS. EFI. NMS. DST. DHS.	The runners perceived the park environment as more restorative (PRS) ($p<0.0001$) than the urban environment. No significant differences were found in mood or attention restoration between time and environment.
Geniole et al. 2016; Canada55	Cross-sectional; $\mathrm{P}=31$; M: 31, F: 0; $\text{mean age}=$ $24.61\pm 3.88$ y; adult males.	Naturalized landfill (designed to resemble a natural and green park) and urban area (commercial and business zone).	Walking.	Affect grid. [mood and attention (arousal)] Stroop test.	Walking in a naturalized landfill increased mood ($\mathrm{t}30=2.88$, $p=0.007$, Cohen’s $d=0.59$) compared with the urban environment ($\mathrm{t}30=0.21$, $p=0.84$, Cohen’s $d=0.05$), but not arousal between locations ($p=0.32$) nor attentional restoration (Stroop test, $p>.64$).
Han 2017; Taiwan30	Cross-sectional; $\mathrm{P}=116$; M: 52, F: 64; $\text{mean age}=$ 20.85 y; university students.	University campus: natural environment (high visible greenness rate) and built environment (low visible greenness rate).	Walking and jogging.	POMS-SF. TMD.	Significant differences were observed for the natural compared with the built environment for two variables of POMS: fatigue (natural: $\text{mean}=6.836$, built: $\text{mean}=8.941$; $p=0.019$) and nervousness (natural: $\text{mean}=4.284$, built: $\text{mean}=5.471$; $p=0.011$). No significant difference was found between the natural and built environment for the TMD measure.
Han 2021; Taiwan31	Cross-sectional; $\mathrm{P}=95$; M: 41, F: 54; $\text{mean age}=$ $20.47\pm 1.2$; university students.	Woodland park.	Walking and jogging.	POMS. TMD.	The main effects of the test time for several measures (POMS and TMD) were significant (F(1,89) $\ge 5.704$, $p\le 0.019$, ${\eta }_p^2\ge 0.060$).
Holt et al. 2019; USA53	Cross-sectional; $\mathrm{P}=207$; H: 31.4%, M: 69.6%; age: ND; university students.	University campus featuring a combination of open green spaces and wooded areas.	Hiking/walking, running, cycling.	PSS.	Active use of green spaces under perceived stress ($p<0.01$).
Johansson et al. 2011; Sweden49	Cross-sectional; $\mathrm{P}=20$; M:10, F: 10; women’s $\text{mean age}=$ $22.4\pm 3.03$ y, men’s $\text{mean age}=$ $24.2\pm 2.81$ y; range 20–29 y; university students.	Municipal park and urban center.	Walking.	EFI. NMS. SST.	While walking through the park, revitalization (positive affect) increased to a greater extent when alone compared with being accompanied ($p\le 0.019$). There was no difference in other mood measures between the park and urban settings.
Kajosaari and Pasanen 2021; Finland45	Cross-sectional; $\mathrm{P}=760$; M: ND, F: ND; range: 18–65 y; adult population.	Blue spaces, outdoor sports facilities, maintained UGS, small urban forests, large urban forests, large recreational forests, built outdoor environments.	Walking, running, or swimming.	Perceived restorative outcomes using a binary scale: Relaxation: “settings that aid relaxation.” Stress reduction: “settings that help escape from stress.”	Relaxation in blue spaces ($\text{OR}=2.51$; $p>0.001$), large urban forests ($\text{OR}=2.11$; $p>0.004$), and large recreational forests ($\text{OR}=3.12$; $p>0.001$). Stress reduction in blue spaces ($\text{OR}=2.78$; $p<0.001$), large urban forests ($\text{OR}=2.12$; $p>0.005$), and large recreational forests ($\text{OR}=3.28$; $p>0.001$).
Kinnafick and Thøgersen-Ntoumani 2014 (1); UK42	Cross-sectional; $\mathrm{P}=40$; M: 8, F: 32; $\text{mean age}=$ $23\pm 7.65$ y; range: 18–60 y; university staff and students.	Laboratory images of natural environment and urban setting.	Walking.	PANAS: arousal. FAS.	Positive affect increased in both natural and urban environments ($p<0.01$), whereas negative affect decreased when observing the natural environment ($p<0.01$). Perceived levels of activation were higher in the natural environment ($p<0.01$).
Kinnafick and Thøgersen-Ntoumani 2014; UK42	Cross-sectional; $\mathrm{P}=30$; M: 13, F: 17; $\text{mean age}=$ of $25.86\pm 11.52$ y; range: 18–61 y; university staff and students.	Real natural environment and built urban environment.	Walking.	PANAS. FAS.	There were no significant differences between natural and urban environments. Perceived levels of activation were higher in the natural environment ($p<0.02$).
Lee et al. 2021; USA54	Cross-sectional; $\mathrm{P}=19$; M: 68%, F: 32%; $\text{mean age}=$ 37 y; recreational adult runners.	Natural environment (public park with natural area and numerous circular trails immersing individuals in a space) and built urban environment (buildings, vehicles, traffic lights, and pedestrians).	Running.	EEG measurements to measure cortical activity corresponding to discrete psychological states (focus, anxiety, internal attention, motivation, relaxation).	Focus decreased immediately from baseline ($\text{mean}=1.307$) to urban race ($\text{mean}=0.845$), slightly increased during the nature part ($\text{mean}=0.937$), and slightly increased in the post-race stage ($\text{mean}=0.948$). Anxiety decreased immediately from baseline ($\text{mean}=0.707$) to urban race ($\text{mean}=0.109$), slightly increased during the nature part ($\text{mean}=0.158$), and slightly increased in the post-race stage ($\text{mean}=0.194$). Internal attention and relaxation followed a similar trajectory to focus and anxiety, with a slight decrease in the final stage. Motivation showed a linear decrease throughout the course.
Marselle et al. 2013; UK43	Cross-sectional; $\mathrm{P}=708$; M: 38%, F: 62%; $\text{mean age}=$ $\ge 55$ y; adult population.	UGS, natural and semi-natural areas (rural park, nature reserve), green corridor (river trail, cycle paths, bridleways), farmland, coastal (sea, estuary), urban public space (streets, shopping centers, square), mixed (“a combination of all of the above”).	Walking.	PSS. PANAS.	Group walks in green corridors ($\mathrm{\beta }=-0.20$, $p=0.005$) and agricultural land environments ($\mathrm{\beta }=-0.17$, $p=0.006$) were associated with significantly lower perceived stress compared with urban public space. There was no significant difference in arousal between green environments and urban public space.
Martens et al. 2011; Switzerland50	Cross-sectional; $\mathrm{P}=96$; M: 56%, F: 44%; $\text{mean age}=$ 37.6 y; students, individuals with regular full-time employment, and older adults.	Wild forest and well-maintained forest.	Walking.	Psychological well-being: self-rating scales of mental state. Positive affect: “good mood” and “calmness.” Negative affect: “anger” and “depression.” Activation: “activation” and “lethargy.” Arousal: “arousal”.	Increased positive affect and positive mood ($p<0.05$) in the well-maintained forest. There was no significant difference in activation and arousal.
Mokhtar et al. 2018; Malaysia27	Cross-sectional; $\mathrm{P}=20$; M: 20, F: 0; $\text{mean age}=$ 21.1 y; male university students.	Urban park and urban environment.	Walking.	ROS. POMS. TMD. PANAS.	ROS—greater restoration in UGS than in the city ($p<0.001$). POMS—greater mood in UGS than in the city for the six scales (tension: UGS: $2.2\pm 0.97$; city: $12.3\pm 2.40$; $p<0.05$, depression: UGS: $2.8\pm 1.10$; city: $10.3\pm 2.18$; $p<0.05$, anger—UGS: $2\pm 0.99$; city: $10.45\pm 2.27$; $p<0.05$, fatigue: UGS: $4.1\pm 1.13$; city: $12.25\pm 1.76$; $p<0.05$, Confusion: UGS: $3.35\pm 1.01$; city: $8.55\pm 1.64$; $p<0.05$), and vigor UGS: $23.55\pm 1.43$; city: $13.65\pm 1.76$; $p<0.05$). TMD—greater between UGS and city (UGS: $-9.71\pm 5.4$; city: $40.2\pm 10.87$; $p<0.05$). PANAS—positive attitudes increased in UGS ($p<0.01$). There was no significant difference in negative attitudes in UGS.
Pasanen et al. 2018; Finland47	Cross-sectional; $\mathrm{P}=128$; M: 20%, F: 80%; $\text{mean age}=$ 50 y; adult population.	Natural trail.	Walking.	ROS. Two-dimensional affect grid to measure mood (activation and valence).	Greater attention restoration with and without cognitive task, ranging from 0.48 to 0.67 units ($p<0.000$). Activation with and without cognitive task, ranging from 1.27 to 2.16 units more tranquil ($p<0.000$). Valence without cognitive task, 1.52 units more pleasant ($p<0.01$).
Pasanen et al. 2018; Finland47	Cross-sectional; $\mathrm{P}=128$; M: 20%, F: 80%; $\text{mean age}=$ 50 y; adult population.	Urban Park.	Walking.	ROS. Two-dimensional affect grid to measure mood (activation and valence).	Greater attention restoration with and without cognitive task, ranging from 0.63 to 0.84 units ($p<0.000$). Valence with and without cognitive task, ranging from 1.17 to 1.66 units more pleasant ($p<0.000$). Activation with cognitive task, $-0.64$ units calmer ($p<0.03$).
Paul et al. 2020; Bangladesh38	Cross-sectional; $\mathrm{P}=181$; M: ND, F: ND; $\text{mean age}=$ 47 y; adult population.	Open public spaces.	Walking/jogging, cycling, exercising, socializing with friends, relaxing, accompanying children, playing (soccer, skating).	Structured questionnaire that included a variety of well-being considerations.	Perception regarding improvement in psychological health (help in reducing stress: ${\mathrm{\chi }}^2=147.39$; $p<0.001$).
Pretty et al. 2007; UK44	Cross-sectional; $\mathrm{P}=263$; M: ND, F: ND; $\text{mean age}=$ $47.8\pm 18.2$ y; adult population.	Different types of green spaces environments (woods with open spaces; water elements, such as lakes or canal sides).	Walking, cycling, horseback riding, fishing, sailing/boating, forest activities, conservation activities.	RSE. POMS.	Engaging in exercise in nature, regardless of type, duration, or intensity, led to significant improvements in self-esteem (RSE) ($p<0.0005$), mood enhancement (POMS), with reduction in anger/hostility ($p<0.0005$), confusion/bewilderment ($p<0.0005$), depression/dejection ($p<0.0005$), and tension/anxiety ($p<0.0005$), as well as fatigue/inertia ($p<0.05$), but not in vigor/activity ($p<\text{NA}$), and improvements in TMD ($p<0.0005$).
Rajoo et al. 2021; Malaysia28	Cross-sectional; $\mathrm{P}=30$; M: 20, F: 10; $\text{mean age}=$ $26.2\pm 4.14$ y; young adult population.	UGS.	Circuit training.	DASS-21: stress.	Stress levels significantly decreased after the nature exercise program ($\text{mean}=13.13$, $\text{sd}=\pm 3.96$), $\mathrm{t}\left(14\right)=3.2$, $p<0.001$.
Raman et al. 2021; Malaysia29	Cross-sectional; $\mathrm{P}=80$; H: ND, M: ND; $\text{mean age}=$ $33.68\pm 12.06$ y. working adult population in the service sector.	Wooded area. Constructed urban park. Urban green corridor.	Walking.	DASS-21: stress. POMS. TMD. PANAS. ROS.	DASS—stress significantly decreased in all three environments: urban green corridor (A: $22.95\pm 8.28$, B: $20.58\pm 9.75$; $p<0.001$), constructed urban park (A: $22.48\pm 8.37$, B: $17.6\pm 10.20$; $p<0.001$), wooded area (A: $22.38\pm 8.36$, B: $10.38\pm 7.71$; $p<0.001$). POMS—significant mood improvement in all three environments, except for vigor in the urban green corridor. TMD—significant improvement in all three environments (urban green corridor: $p<0.05$; constructed urban park: $p<0.1$; wooded area: $p<0.1$). PANAS—positive affect increased more in the wooded area (B: $33.69\pm 0.65$; A: $46.49\pm 0.35$; $p<0.05$) and the constructed urban park (B: $23.93\pm 0.71$; A: $33.80\pm 0.75$; $p<0.05$) compared with the urban green corridor (B: $24.36\pm 0.94$; A: $28.64\pm 0.89$; $p<0.05$). Meanwhile, negative affect decreased rapidly in the wooded area (B: $17.75\pm 0.65$; A: $5.59\pm 0.33$; $p<0.05$) compared with the other two experimental environments. ROS—improvement in attention restoration in all three environments ($p<0.001$).
Song et al. 2013; Japan33	Cross-sectional; $\mathrm{P}=13$; M: 13, F: 0; $\text{mean age}=$ $22.5\pm 3.1$ y; male university students.	Urban park and city area.	Walking.	POMS.	The score of the negative subscale tension/anxiety was significantly lower after walking in the urban park in winter compared with the city area ($p<0.01$). Conversely, the positive mood for vigor was significantly higher in the urban park in winter but not in the city area ($p<0.01$).
Song et al. 2014; Japan34	Cross-sectional; $\mathrm{P}=17$; M: 17, F: 0; $\text{mean age}=$ $21\pm 1.7$ y; male university students.	Urban park and city area.	Walking.	POMS.	The negative subscales of tension/anxiety and fatigue were significantly lower after walking in the urban park in spring than after walking in the city area ($p<0.05$). Conversely, the positive mood state vigor was significantly higher after the walk in the urban park in spring ($p<0.05$).
Song et al. 2015; Japan35	Cross-sectional; $\mathrm{P}=23$; M: 23, F: 0; $\text{mean age}=$ $22.3\pm 1.2$ y; male university students.	Urban park and city area.	Walking.	POMS.	Scores for the negative subscales of tension/anxiety, anger/hostility, fatigue, and confusion were significantly lower after walking in the urban park in fall than in the city area ($p<0.05$). Conversely, the positive mood vigor was significantly higher after walking in the urban park ($p<0.01$.
Mental well-being
Flowers et al. 2022; Australia56	Cross-sectional; $\mathrm{P}=22$; M: 11, F: 11; $\text{mean age}=$ 24.4 y; range: 20 to 32 y; university students.	Green natural environment and urban environment.	Walking.	STAI.	There was no evidence of an overall effect of walking environment (green vs. urban) on the change in anxiety ($\mathrm{\beta }=2.81$, 95% CI: $-1.83$ to 7.47, $p=0.22$).
Marselle et al. 2013; UK43	Cross-sectional; $\mathrm{P}=708$; M: 38%, F: 62%; $\text{mean age}=$ $\ge 55$ y; adult population.	UGS, natural and semi-natural areas (rural park, nature reserve), green corridor (river trail, cycle paths, bridleways), farmland, coastal (sea, estuary), urban public space (streets, shopping centers, square), mixed (“a combination of all of the above”).	Walking.	MDI.	There were no significant differences between group walks in urban public space and walks in any type of natural environment.
Rajoo et al. 2021; Malaysia28	Cross-sectional; $\mathrm{P}=30$; M: 20, F: 10; $\text{mean age}=$ $26.2\pm 4.14$ y; young adult population.	UGS.	Circuit training.	DASS-21: depression and anxiety.	Anxiety significantly decreased after the nature exercise program ($\text{mean}=5.27$, $\text{sd}=\pm 3.6$), ($\mathrm{t}\left(14\right)=3.1$, $p<0.02$). Depression significantly decreased after the nature exercise program ($\text{mean}=7.2$, $\text{sd}=\pm 7.2$), ($\mathrm{t}\left(14\right)=3.0$, $p<0.01$).
Raman et al. 2021; Malaysia29	Cross-sectional; $\mathrm{P}=80$; H: ND, M: ND; $\text{mean age}=$ $33.68\pm 12.06$ y. working adult population in the service sector.	Wooded area, constructed urban park, urban green corridor.	Walking.	DASS-21: depression and anxiety.	Anxiety significantly decreased in all three environments: urban green corridor (A: $18.4\pm 10.65$, B: $20.58\pm 16.38$; $p<0.001$), constructed urban park (A: $19.2\pm 10.15$, B: $15.4\pm 10.76$; $p<0.001$), wooded area (A: $17.98\pm 10.00$, B: $8.08\pm 7.34$; $p<0.001$). Depression significantly decreased in all three environments: urban green corridor (A: $22.83\pm 8.04$, B: $20.28\pm 8.62$; $p<0.001$), constructed urban park (A: $17.85\pm 10.40$, B: $14.58\pm 10.75$; $p<0.001$), wooded area (A: $16.93\pm 10.96$, B: $7.55\pm 8.06$; $p<0.001$).
Shanahan et al. 2016; Australia57	Cross-sectional; $\mathrm{P}=\mathrm{1,538}$; M: ND, F: ND; range: 18–65 y; adult population.	Public green spaces.	Walking.	DASS: depression.	An inverse correlation between physical activity frequency and depression is observed, as indicated by a pseudo $R^2$ of $-0.13$ ($p<0.03$), suggesting that as physical activity frequency increases, depression tends to decrease.
Song et al. 2013; Japan33	Cross-sectional; $\mathrm{P}=13$; M: 13, F: 0; $\text{mean age}=$ of $22.5\pm 3.1$ y; male university students.	Urban park and city area.	Walking.	STAI.	The anxiety score was 18.2% lower in the urban park in winter ($37.3\pm 8.7$ scores) compared with the city area ($45.6\pm 7.1$ scores; $p<0.05$).
Song et al. 2014; Japan34	Cross-sectional; $\mathrm{P}=17$; M: 17, F: 0; $\text{mean age}=$ $21\pm 1.7$ y; male university students.	Urban park and city area.	Walking.	STAI.	The anxiety score was 14.3% lower after walking in the urban park in spring compared with walking in the city area (UP: $41.6\pm 7.0$, CA: $48.6\pm 6.3$; $p<0.05$).
Song et al. 2015; Japan35	Cross-sectional; $\mathrm{P}=23$; M: 23, F: 0; $\text{mean age}=$ $22.3\pm 1.2$ y; male university students.	Urban park and city area.	Walking.	STAI.	The anxiety score was 19.3% significantly lower after walking in the urban park in fall than in the city area (UP: $39.0\pm 6.3$; CA: $48.4\pm 7.5$; $p<0.01$).
Social well-being
Hakman et al. 2021; Ukraine51	Cross-sectional; $\mathrm{P}=101$; M: 50, F: 51; range: 60–65 y; adult population.	Urban park.	Physical exercise sessions (warm-up: stretches, main part: aerobic exercises (walking + strength exercises), cool-down: breathing exercises + stretches).	SF-36: social activity.	High levels of social activity were observed post-intervention ($72.3\pm 0.7$ points in men and $78.5\pm 0.9$ points in women).
Paul et al. 2020; Bangladesh38	Cross-sectional; $\mathrm{P}=181$; M: ND, F: ND; $\text{mean age}=$ 47 y; adult population.	Open public spaces.	Walking/jogging, cycling, exercising, socializing with friends, relaxing, accompanying children, playing (soccer, skating).	Structured questionnaire that included a variety of well-being considerations.	Perception of improvement in social well-being (Feeling supportive toward others: ${\mathrm{\chi }}^2=214.95$; $p<0.001$, Assistance in making friendships: ${\mathrm{\chi }}^2=101.27$; $p<0.001$).
Shanahan et al. 2016; Australia57	Cross-sectional; $\mathrm{P}=\mathrm{1,538}$; M: ND, F: ND; range: 18–65 years of age; adult population.	Public green spaces.	Walking.	The respondents’ perceptions of social cohesion were estimated based on three previously developed questions measuring trust, reciprocal exchange within communities, and overall community cohesion.	Higher frequency of physical activity correlates with greater social cohesion, with an $R^2$ of 0.03 ($p<0.01$).

Note: A, after; Affect grid (measures positive mood and arousal); B, before; BDST, backward digit span test (measures concentration); CA, city area; Chester Step Test (measures aerobic fitness); CI, confidence interval; DASS, Depression, Anxiety, and Stress Scale; DASS-21, Depression, Anxiety, and Stress Scale (with 21 items); DBP, diastolic blood pressure; DHS, Daily Hassles Scale [to measure the need for restoration (stress)]; DST, Digit span test [consists of digit span forward and digit span backward tests and the Symbol Digits Modalities Test (measures attention)]; DSB, digit span backward; DSF, digit span forward; EEG, electroencephalogram; EFI, Exercise-Induced Feeling Inventory (measures positive emotion); F, female; FAS, Felt Arousal Scale (measures arousal levels); HF, high frequency; HR, heart rate; HRV, heart rate variability; M, male; M, mean; MDI, Major Depressive Inventory (measures depression); ND, no data available; NMS, negative mood scale (measures negative emotion); OR, odds ratio; P, participants; PANAS, Positive and Negative Affect Schedule (measures positive and negative effects); POMS, Profile of Mood States test (measures mood); POMS-SF, Profile of Mood State short form; PRS, Perceived Restorativeness Scale (measures restoration); PSS, Perceived Stress Scale (measures stress); PWC150 test, aerobic power index test (measures physical performance); RMSSD/rMSSD, root mean square of successive differences; ROS, Restoration Outcome Scale (measures restoration); RSE, Rosenberg self-esteem scale (measures self-esteem); SBP, systolic blood pressure; SD, standard deviation; SD1, measures rapid changes in R-R interval; SDNN, standard deviation of RR interval; SF-36, Medical Outcomes Study Questionnaire Short Form (measures social activity); SF-8, health questionnaire (measures participants’ perceived general and physical health); SST, Symbol Substitution Test (measures attention); STAI, State-Trait Anxiety Inventory (measures anxiety state); Stroop test (measures concentration level); TMD, Total Mood Disturbance (measures overall mood disturbance); UGS, urban green space; UP, urban park; ${\mathrm{\chi }}^2$, chi-squared.

Activities and Green Spaces

All included studies were conducted to assess the influence of engaging in physical and sporting activities in urban natural environments. The activities performed were diverse; although walking predominated ($n=20$), there were also, to a lesser extent, other activities, such as running ($n=3$), or studies that included both running and walking ($n=2$), cycling ($n=1$), physical exercise sessions ($n=2$), and swimming ($n=1$), as well as observational studies with different types of physical exercises ($n=2$). These activities took place in various urban green spaces, depending on each author and their definition of green space, including urban parks, urban forests, forest parks, university campuses, naturalized landfills, natural and heritage green spaces, public green spaces, natural environments, different green spaces and natural settings, natural trails, and open public spaces; descriptive information about these spaces was detailed in the majority of the documents (Table 2).

Types of Studies

Eighteen studies compared two different spaces. Physical exercise intervention in urban green space was compared with that in an urban environment (referring to urbanized areas with minimal or no vegetation, including densely populated residential areas, commercial or industrial zones, and city streets with no significant green spaces),^{27,30,33–36,40–43,46,48,49,54–56} one study compared an urban green space with an indoor space,⁵² and one study compared a forest park with an urban park.³² The remaining studies focused on physical activity in a single urban green space.

Benefits of Exercise in Urban Green Spaces

The main results of each study are presented in Table 2. The outcomes considered in this systematic review were organized into four components: physical, psychological, mental, and social well-being.

Physical Well-Being

Fifteen articles assessed physical well-being,^{27,32–38,40,41,46,51,53,55,57} encompassing direct measures through physiological metrics and indirect measures using questionnaires or physical tests. Among the physiological measures, heart rate (HR) was the most used (8/15, 53%),^{27,32–36,40,46} followed by heart rate variability (HRV) (7/15, 46%),^{33–36,40,41,46} blood pressure (5/15, 33.3%),^{27,32,40,46,57} salivary cortisol (2/15, 13.3%),^27,55 testosterone (1/15, 6.6%),⁵⁵ vascular function (1/15, 6.6%)³² and cardiac function [left ventricular pressure slope (LV dp/dt maximum) and cardiac output] (1/15, 6.6%).³² Among the indirect measures, the self-rated general health questionnaire was the most used (3/15, 20%)^40,51,53 along with self-rated measures of physical health (3/15, 20%),^37,38,51 followed by cardiovascular fitness (2/15, 13.3%),^40,51 physical activity level (1/15, 6.6%),⁵¹ cardiovascular risk (1/15, 6.6%)⁴⁰ and body mass index (BMI) (1/15, 6.6%).⁴⁰

The results obtained for physical well-being measures were diverse. Seven of eight articles observed that walking in urban green spaces was associated with significantly reduced HR compared with walking in an urban environment, suggesting greater relaxation,^27,32,46 regardless of the season.^33–36 However, Brown et al.⁴⁰ found no significant differences in HR between the group walking in green environments, the urban environment group, and the waiting control group, who did not participate in any intervention during the study. Five of seven studies found that individuals had significantly greater HRV when walking in green spaces compared with when walking in urban environments, associated with outcomes related to a better relaxation response^33–36 and measures of more restorative sleep.⁴¹ However, Lanki et al.⁴⁶ found no significant differences in HRV between walking in an urban forest, an urban park, and the city center. Brown et al.⁴⁰ found no significant HRV differences between walking in a natural environment, walking in an urban environment, and the control group.

Diastolic blood pressure (DBP) was observed to be significantly lower in individuals exercising in natural environments compared with those in built environments.^27,40,46 However, associations with systolic blood pressure (SBP) were less consistent, with some studies showing nonsignificant differences,⁴⁶ whereas others found no significant differences between those exercising in natural and built environments.²⁷ One study found no association between the frequency of exercise in urban green spaces and improvements in blood pressure.⁵⁷ Another study comparing a walk in a natural environment and a walk in a park found no significant differences, with both green environments being associated with favorable blood pressure outcomes.³²

Regarding salivary cortisol, the results of the few available studies were contradictory. Although Mokhtar et al.,²⁷ in a cross-sectional study with 20 male university students, found significantly lower salivary cortisol in participants when walking in an urban green space compared with those walking in an urban environment, Geniole et al.,⁵⁵ in a cross-sectional study with 31 adult males, found no significant differences in salivary cortisol between participants walking in a naturalized space and those walking in an urban environment. There were also no significant differences in testosterone levels between participants walking in a naturalized space and those walking in an urban environment.⁵⁵ Tsao et al.³² found significant differences between walking in an urban forest and in an urban park in terms of cardiac function and vascular function, with the greener and more natural environment being associated with beneficial outcomes for cardiovascular health.

Self-rated physical health was measured with different instruments for indirect measures of physical well-being. Hakman et al.⁵¹ and Brown et al.⁴⁰ used the SF-36 short version and the SF-8 questionnaire, respectively, whereas Holt et al.⁵³ used a Likert-scale question on self-rated general health. It was found that the active use of green spaces by walking, running, or cycling was associated with better general health⁵³ and that exercising in a park was associated with higher scores for general health.⁵¹ However, Brown et al.⁴⁰ found no significant differences in self-rated general health between those walking in a natural environment, an urban environment, and the control group. For physical health, Hakman et al.⁵¹ found that exercising in an urban park was associated with higher physical health scores measured at baseline (prior to starting the intervention) and again at the end of the intervention. Paul et al.,³⁸ in an observational and cross-sectional study using a nonvalidated structured questionnaire, found that exercising in open public spaces was associated with better physical health according to users’ perceptions in several measures: They found it helped reduce or maintain body weight, reduce or maintain hypertension, lower blood sugar levels, and reduce fatty liver. On the other hand, Mao et al.³⁷ measured physical health through a physical evaluation test [National Student Physical Health Standard of China (NSPHS)] in university students. A longitudinal study that evaluated 2,852 students before and after a running program on a university campus with vegetation and trees showed a significant improvement in test results after the intervention. Hakman et al.⁵¹ also found that exercising in urban parks was associated with improved physical activity levels and cardiovascular fitness. However, Brown et al.⁴⁰ found no significant differences in cardiovascular fitness, cardiovascular risk, and BMI between those walking in natural spaces and those walking in urban environments or the control group.

Psychological Well-Being

A total of 22 articles,^{27–31,33–35,38,39,42–45,47–50,52–55} were selected to evaluate mental well-being, including 2 that contained 2 studies each.^42,47 Among the most evaluated measures were mood (16/22, 72.7%),^{27,29–31,33–35,39,42–44,47–50,55} followed by restoration (7/22, 13.6%),^{27,29,47–49,52,55} stress (6/22, 27.7%),^{28,29,38,43,45,53} self-esteem (2/22, 9%),^39,44 arousal (2/22, 9%),^45,55 and meditative states through electroencephalogram (EEG; 2/22, 9%).^52,54 There was a discrepancy between the total number of psychological well-being outcomes assessed and the total number of included studies because most studies evaluated multiple outcomes simultaneously.

The main results revealed that 13 of the 16 articles found significant associations between exercising in urban green spaces and improved mood,^{27,29–31,39,42,44,47,50,55} even in different seasons of the year.^33–35 However, Bodin and Hartig,⁴⁸ in a cross-sectional study, found no significant associations between running in an urban green space and mood compared with running in an urban environment. Similarly, Marselle et al.⁴³ also found no significant associations in mood when walking in different types of green spaces.

Other studies revealed partial results, such as Johansson et al.,⁴⁹ who found significant associations in the improvement of positive mood effects, such as vitality, energy, and a sense of revitalization, but not in negative mood effects, such as sadness, anxiety, and irritability. On the other hand, Kinnafick and Thøgersen-Ntoumani⁴² found significant associations in mood measured in the laboratory while participants walked observing images of nature on a screen but found no significant differences when walking in a natural environment compared with an urban environment. Restoration was evaluated from three different perspectives: a) attention restoration, which referred to the process whereby a person’s attention capacity recovers after being fatigued; b) mental restoration, which encompassed a broader concept including the recovery of cognitive and emotional capacity; and c) self-reported restoration, which referred to a person’s subjective perception of their state of recovery.

Of the four studies that measured attention restoration, only the study by Bailey et al.⁵² found significant differences in the Stroop test after walking in a green space compared with walking in an indoor environment. However, no significant differences were found for the backward digit span test (BDST). No significant differences were found in the other studies on attention restoration.^48,49,55 Regarding mental restoration, only one study, by Bodin and Hartig,⁴⁸ found a significant difference in mental restoration using the Perceived Restorativeness Scale (PRS) in participants running in a natural park compared with those running in an urban environment. Three studies found higher measures of self-reported restoration when walking in urban green spaces, both in comparative²⁷ and noncomparative studies.^29,47

Six reviewed studies found a significant difference in stress reduction when exercising in urban green spaces in noncomparative studies,^{28,38,43,45,53} as did one comparative study.²⁹ Conversely, in two studies,^39,44 self-esteem was higher after exercising in different types of green environments, as measured by the Rosenberg self-esteem scale (RSE).

Arousal, which evaluates the general activation and excitement of the nervous and emotional system, was also measured. The results were contradictory: Kinnafick and Thøgersen-Ntoumani⁴² found that perceived arousal levels were higher in natural environments, both in the laboratory observing images of nature and when walking in a natural environment compared with an urban environment. However, Geniole et al.⁵⁵ found no significant differences for arousal between walking in a naturalized environment and walking in an urban environment.

Meditative states were measured in two studies^52,54 using EEG. Bailey et al.⁵² found that walking in a green space significantly improved states of meditation and relaxation compared with walking in an indoor space. Similarly, Lee et al.⁵⁴ found that running in a natural environment compared with running in an urban environment improved meditation, relaxation, and anxiety.

Mental Well-Being

Using validated questionnaires, eight articles were considered to measure mental well-being.^{28,29,33–35,43,56,57} The evaluated measures included depression (4/8, 50%) assessed through the Major Depressive Inventory (MDI)⁴³ and the depression, anxiety, and stress scale (DASS)^28,29,57 questionnaires, and anxiety (6/8, 75%) assessed through the State-Trait Anxiety Inventory (STAI)^33–35,56 and DASS^28,29 questionnaires. The total number of mental well-being outcomes evaluated was greater than the total number of included studies because most evaluated more than one outcome simultaneously.

The main results for mental well-being measures were diverse. Three of the four articles found that exercising in urban green spaces was associated with lower depression levels.^28,29,57 However, in an observational study, Marselle et al.⁴³ found no significant differences in depression measures for people walking in urban green spaces. On the other hand, five of the six articles found an association between exercising in urban green spaces and anxiety reduction, both in noncomparative studies^28,29 and in comparative studies between an urban green space and an urban environment across different seasons of the year.^33–35 Nevertheless, Flowers et al.⁵⁶ found no significant evidence of anxiety reduction in a cross-sectional study comparing 11 women and 11 men walking in an urban green space and an urban environment.

Social Well-Being

Three articles were considered to measure social well-being^38,51,57 using questionnaires. The evaluated measures included social interaction (2/3, 66.6%), measured through the validated SF-36 questionnaire in its social activity dimension⁵¹; social cohesion (2/3, 66.6%), assessed based on three previously developed questions addressing trust, reciprocal exchange within communities, and general community cohesion⁵⁷; and a nonvalidated structured questionnaire that explored users’ perceptions of social well-being considerations, such as social interaction and social cohesion.³⁸ The total number of social well-being outcomes evaluated was greater than the total number of included studies because most of them evaluated more than one outcome simultaneously.

The main results indicated that exercising in urban green spaces was associated with better social interaction^38,51 and social cohesion.^38,57 These findings underscore the essential role of urban green spaces as facilitators of interaction among individuals and as promoters of the formation of healthy social relationships.

Relationship between Type of Physical Activity and Well-Being

In addition to the primary analyses, a complementary analysis was conducted to examine the relationship between the type of physical activity in urban green spaces and various dimensions of well-being (physical, psychological, mental, and social). Table 3 summarizes these results, where the symbols indicate whether the relationship was positive or nonsignificant, and the numbers in parentheses represent the number of studies supporting each relationship.

Table 3.

Relationship between type of physical activity and well-being.

Physical activity	Physical well-being	Psychological well-being	Mental well-being	Social well-being
Walking	$+\left(8\right)$	$+\left(11\right)$	$+\left(6\right)$	$+\left(3\right)$
Running	$+\left(3\right)$	$+\left(5\right)$	$-\left(1\right)$	$+\left(1\right)$
Physical exercise	$+\left(2\right)$	$-\left(1\right)$	$-\left(1\right)$	$+\left(1\right)$
Walking, jogging, cyclinga	$+\left(2\right)$	$+\left(2\right)$	$-\left(1\right)$	$+\left(1\right)$
Swimming	—	$+\left(1\right)$	—	—
Circuit training	—	$+\left(1\right)$	$+\left(1\right)$	—

Note: The “+” sign indicates a statistically significant positive relationship in that well-being category, while “–” indicates a nonsignificant relationship. Numbers in parentheses represent the number of studies documenting that specific relationship. —, no relevant information was found for that combination of activity and well-being category.

^a These exercises appeared equally in the study and cannot be differentiated from each other.

As shown in the table, walking was the most studied type of physical activity, demonstrating a positive association with physical, psychological, mental, and social well-being in several studies ($n=8$ for physical well-being, $n=11$ for psychological well-being). Other activities, such as running and circuit training, also showed significant associations, although in a smaller number of studies.

The results suggest that physical activity in urban green spaces can have a notable positive impact on various dimensions of well-being, especially physical and psychological well-being, regardless of the type of activity. However, some activities, such as swimming, show limited evidence, indicating the need for further research in this area.

Discussion

The systematic review aimed to analyze the scientific literature investigating the relationship between physical and sporting activities in urban green spaces and the well-being of the individuals participating in them. We identified 31 studies that met our inclusion criteria and reported a positive association between green exercise and physical, psychological, mental, and social well-being. Some of these studies compared the effects of outdoor activities in green spaces with those conducted in other urban or indoor environments.

Our findings revealed positive associations between exercising in natural environments and various measures of physical well-being, such as HR, HRV, and DBP. To a lesser extent, published studies also found positive associations with cardiac function and vascular function, although these latter findings were based on a smaller number of studies. Other measures, such as salivary cortisol, self-rated health, and physical condition, showed less consistent results owing to the limited amount of available research.

These results contrast with the systematic review by Lahart et al.,¹⁸ which did not find positive associations between exercising in natural environments and measures of HR and DBP compared with indoor exercise. Similarly, Coventry et al.,⁵⁸ did not find a positive relationship between exercising in natural environments and DBP. These discrepancies may be due to differences in study design, populations investigated, or methodologies used. In the context of cardiovascular health, the study by Bauer and White⁵⁹ also emphasized the benefits of spending time in nature, especially in the prevention and treatment of hypertension. Their results suggest that activity in urban green spaces can have a positive impact on blood pressure and reduce the need for antihypertensive medication. This further highlights the importance of considering outdoor physical activity in urban green spaces as part of prevention and treatment strategies.

The findings of this systematic review support the idea that physical activity in urban green spaces not only can offer significant benefits to physical well-being but also has the potential to positively impact psychological and mental well-being. Participants engaging in physical activity in urban green spaces experienced better mood, stress reduction, self-esteem, and perceived restoration, although the results are less consistent for mental restoration. These benefits were observed in studies with varied comparison groups: Some compared green spaces to built environments, whereas others did not include a comparative group. In addition, lower levels of depression and anxiety were observed, suggesting a preventive effect on mental health. These findings align with the results obtained by the systematic review by Ballester-Martínez et al.,¹¹ who suggested that engaging in green exercise compared with indoor exercise is associated with psychological benefits, such as increased positive affect and decreased negative affect, which are related to better mood and physiological stress responses. Our results also align with another systematic review and meta-analysis by Wicks et al.,⁶⁰ which suggested that activity in urban green spaces compared with exercise in urban environments has beneficial effects on emotions, such as positive and negative affect (mood), anxiety, and, with somewhat weaker evidence, depression.

In line with the presented evidence, it is important to note that Roberts et al.¹⁵ conducted a systematic review and meta-analysis of evidence on short-term exposure (through active, passive, and mixed participation studies) to the natural environment on depressive mood. Their findings support the positive trend of lower depression related to exposure to the natural environment. In contrast, we found that Lahart et al.¹⁸ did not obtain conclusive results regarding the improvement of mood or depressive symptoms with exercise in natural environments compared with indoor exercise. Similarly, Coventry et al.⁵⁸ did not find a positive association between exercising in natural environments and stress improvement.

The results of our research underscore the increase in social interaction and cohesion within urban green spaces through physical activity, supporting the crucial role of these environments in promoting healthier and more connected communities. However, it is essential to recognize the complexities of factors influencing social dynamics in these settings. For example, Wan et al.,⁶¹ had contrasting findings, suggesting that exercise alone may not substantially improve social cohesion, whereas participation in sedentary activities might. This incongruity highlights the need for further investigation into the nuanced relationship between different forms of activities, environmental perceptions, and social outcomes within urban green spaces. We believe that a comprehensive understanding of these dynamics will be instrumental in designing tailored interventions that effectively foster social well-being in urbanized societies.

Our findings emphasize the importance of noting that, of the 31 articles included in our systematic review, $\sim 56\%$ of the research focused on walking activity. These results align with the analysis by Grigoletto et al.⁶² on outdoor activity in urban green spaces and its impact on adult health. As they suggest, walking activity has been highlighted as one of the most popular practices in these environments and has been shown to be beneficial for health. Walking is an accessible and cost-effective activity that does not require special skills, making it an affordable option for people of all ages. Furthermore, it allows individuals to choose when and where to engage in this activity, promoting an active and healthy lifestyle. Our findings support these results and enhance the understanding of how physical activity in urban green spaces can improve the physical, psychological, mental, and social well-being of participants.

In our analysis of the literature, we observed that, even within the more specific term of “urban green space,” the definition of these spaces can vary considerably depending on the context and discipline. This finding is consistent with the study by Taylor and Hochuli,⁴ which revealed the lack of a unified definition of green space. Given this challenge, we encourage researchers to provide operational and quantifiable definitions of urban green space in their studies, following the recommendation of Taylor and Hochuli.⁴ This practice will contribute to the clarity and comparability of results, fostering a more multidisciplinary and interdisciplinary approach to urban green space research.

This systematic review highlights several significant strengths in our assessment of the relationship between physical activity in urban green spaces and human well-being. First, a diverse set of 31 studies was included, addressing the impact of physical activity in urban green spaces on various dimensions of well-being. This comprehensive inclusion provides a thorough and robust overview of the available evidence. Furthermore, it is important to emphasize the consistency in our findings. The evidence consistently supports the notion that physical activity in urban green spaces has a positive impact on well-being in terms of physical, psychological, mental, and social aspects. This consistency strengthens the validity of our conclusions and underscores the robustness of the observed relationship.

However, our study has certain limitations that warrant consideration. One key limitation is the variability in study quality and the heterogeneity across the included studies, which complicates the ability to draw consistent conclusions. In addition, the presence of a high risk of bias in some studies and the limited quality of available evidence may impact the level of confidence in our findings. These factors suggest the need for caution when interpreting the results and highlight the importance of future research to address these limitations.

Furthermore, the heterogeneity of the studies included in our review poses a significant concern, given that they encompass various methodological approaches, age groups, and geographical contexts. This diversity could potentially affect the generalization of our results. Therefore, caution is advised when extrapolating our conclusions to specific populations or particular contexts. Another critical limitation is the lack of a unified definition of urban green spaces among the included studies, which could hinder the comparison and synthesis of findings. Thus, we advocate for researchers to provide operational and quantifiable definitions of these spaces in future studies to enhance the clarity and comparability of results.

Regarding future directions, we propose several initiatives to advance the understanding of the relationship between physical activity in urban green spaces and human well-being. First, we recommend longitudinal research be conducted to understand how continuous use of these spaces influences well-being over time. In addition, promoting multidisciplinary collaboration is crucial to consider factors such as environmental perceptions, design, and other contextual elements that influence the relationship between activity and well-being. Second, dedicated research is needed in developing countries, particularly those in Latin America and Africa. Studying these regions will allow for the evaluation of the applicability of existing findings in diverse socioeconomic and cultural contexts. It can also help identify unique challenges and opportunities for promoting physical activity in urban green spaces and improving the well-being of local populations. Finally, we recommend studies measure and analyze contextual factors, such as access, quality, and design of urban green spaces, to understand their influence on the well-being of participants. These additional steps will strengthen our understanding of how these environments promote health and well-being globally.

Conclusions

This systematic review aimed to explore the relationship between physical activity in urban green spaces and the well-being of individuals participating in these activities. Our findings provide a comprehensive and robust insight into how physical activity in such environments can impact human well-being across multiple dimensions.

In terms of physical well-being, our results consistently support the notion that physical activity in urban green spaces is associated with improvements in cardiovascular health, such as reductions in blood pressure, HR, and HRV. In addition, positive effects were observed on cardiac and vascular function.

Our findings support benefits in psychological and mental well-being from activity in these urban green environments. Participants experienced better mood, stress reduction, self-esteem, perceived restoration, and, to a lesser extent, mental restoration, as well as meditative states measured with EEG. Furthermore, lower levels of depression and anxiety were associated with green exercise, suggesting a preventive effect on mental health.

Associations with social well-being were also noted, given that urban green spaces are suggested to promote greater social interaction and cohesion, contributing to healthier and more connected communities. This underscores the importance of considering both the physical characteristics and the perceptions and usage patterns of these spaces in future research.

In summary, this systematic review highlights the importance of physical activity in urban green spaces as a promoter of human well-being across its multiple dimensions. We encourage researchers, health professionals, and urban planners to carefully consider the integration of these spaces into public health and urban planning strategies, thus promoting an active and healthy lifestyle that benefits society as a whole.

Conclusions and opinions are those of the individual authors and do not necessarily reflect the policies or views of EHP Publishing or the National Institute of Environmental Health Sciences.