Table 1.

Summary of study characteristics, results and quality appraisal.

Study Reference	Population Description	Sample Size	Cohort Name/Data Source	Follow-Up Duration	Exposure Indicator Description	Exposure Indicator Type	Outcome	Outcome Measure	Main Results Effect Estimate [95% CI 1]	Confounders	Study Quality *
Primary Outcomes
Mental Health
Banay et al. [50]	women nurses; ≥30–55 years [USA]	121,701	Nurses’ Health Study	10 years	NDVI 1 averages for each year of follow-up; 250 m and 1250 m circular buffers	Availability	Depression	First self-report of physician/clinician diagnosis of depression or new regular use of antidepressants	250m Buffer HR 1: 0.87 [0.78, 0.98] Highest NDVI quintile 1250 m Buffer HR: 0.90 [0.80, 1.02] Highest NDVI quintile	age, race, mental health, marital status, educational attainment, husband’s educational attainment, population density, income, median home value, PM 1 2:5 level, BMI 1, smoking status and pack-years of smoking, alcohol consumption, physical activity, physical function, bodily pain [baseline], social network strength, care to ill family members [baseline], difficulty sleeping [baseline]	Poor
Fernandez-Nino et al. [51]	men and women; ≥55 years [Mexico]	1524	Study on Global Ageing and Adult Health [SAGE]	5 years	Street trees; total length of street covered in trees in a 950 m road network buffer	Accessibility	Depression	Self-report of physician diagnosis	OR 1: 0.90 [0.29, 2.83] Highest quintile of street length covered in trees	sex, age, income index, functional limitations, margination index of the municipality	Good
Gariepy et al. [52]	men and women; ≥18–80 years [Canada]	13,618	National Population Health Survey	10 years	Presence of a park within a 500 m circular buffer	Accessibility	Depression	Self-reported instrument	Β 1: −0.4 [−1.4, 0.6] For answering “yes” to presence of a park	age, sex, marital status, education, income adequacy, childhood life events, chronic condition, family history of depression	Good
Gariepy et al. [53]	men and women; ≥18–80 years; with diabetes [any type] [Canada]	2003	Diabetes Health Study [DHS]	5 years	NDVI	Availability	Depression	Self-reported instrument	HR: 0.94 [0.88, 1.01] Per decile increase in NDVI	sex, age, marital status, family income, educational level, employment	Good
Melis et al. [54]	men and women; ≥20–65 years [Italy]	547,263	Turin Longitudinal Study [TLS]	2 years	Availability of green space measured via index by area units	Availability	Depression	Antidepressant use	Men IRR 1: 0.98 [0.92, 1.04] Highest index value quintile green Women IRR: 1.00 [0.96, 1.08] Highest index value quintile of green	sex, age, education level, activity status, citizenship, residential stability at same address	Good
Picavet et al. [55]	men and women; ≥18 to 55 years [Netherlands]	4917	Doetinchem Cohort Study	15 years	Percent green space in 125 m and 1000 m circular buffer	Availability	Depression	Self-reported instruments	Per unit increase in percent green space 125 m OR: 0.97 [0.92, 1.04] 1000 m OR: 0.86 [0.79; 0.93]	age, sex, SES 1	Poor
Tomita et al. [56]	men and women; mean 20 years [South Africa]	11,156	South African National Income Dynamics Study [SA-NIDS	4 years	NDVI, 250 m resolution square	Availability	Depression	Self-reported instrument	OR: 1.01 [1.01, 1.02] Each unit increase in NDVI value	age, sex, marital status, race, household income, employment, rurality	Good
Astell-Burt and Feng [57]	men and women; ≥45 years [Australia]	46,786	45 and Up Study	6.2 [mean] years	Total percent green space; tree canopy in a 1600 m road network buffer	Availability	Depression or anxiety	Self-report of doctor diagnosed	OR: 1.26 [0.89, 1.63] Highest percent quintile total green OR: 0.86 [0.80, 1.01] Highest percent quintile tree canopy	age, sex income, education, economic status, couple status	Poor
Pun et al. [58]	men and women; ≥57–85 years [USA]	3005	National Social Life, Health, and Aging Project [NSHAP]	6 years	NDVI seasonal changes in 1000 m circular buffer	Availability	Depression; anxiety	Self-reported instrument	Anxiety β: −0.104 [−0.322, 0.115] per unit increase in NDVI Depression β: −0.274 [−0.596, 0.048] per unit increase in NDVI	age, gender, questionnaire year, season, region, education attainment, 3-day moving average of temperature, 60-months moving average of PM2.5	Good
Chang et al. [59]	men and women mean age: 43.36 [20.44] years [Taiwan]	869,484	Taiwan Longitudinal Health Insurance Database	10 years	NDVI at baseline; 2000 m circular buffer around hospital most frequently visited	Availability	Schizophrenia	Physician-diagnosed	HR: 0.37 [0.25, 0.55] Highest NDVI quintile	age, sex, health insurance rate, classification of the insured, temperature, relative humidity, precipitation	Good
NCDs
Dalton and Jones [60]	men and women; mean 59.2 years [United Kingdom]	25,639	European Prospective Investigation of Cancer [EPIC] Norfolk	14.5 [mean] years	Percent green space in 800 m circular buffer	Availability	CVD 1	Health register	HR: 0.93 [ 0.88, 0.97] Highest percent quintile green	sex, age, BMI, diabetes, SES [individual and neighbourhood]	Good
Tamosiunas et al. [61]	men and women; ≥45–72 years [Lithuania]	5112	Health, Alcohol, and Psychosocial Factors in Eastern Europe [ HAPIEE]	4.41 [mean] years	Distance to park and park use [self-reported]	Accessibility	CVD	Self-reported doctor diagnosed	User: HR: 1.58 [0.95, 2.63] Longest distance quintile Nonuser: HR: 1.66 [1.01, 2.73] Longest distance quintile	age, sex, education, smoking, arterial hypertension, physical activity, total cholesterol level, fasting glucose level, BMI, diabetes mellitus, cognitive function, symptoms of depression, self-rated health, and quality of life	Good
Clark et al. [62]	men and women; ≥45–84 years; urban residents [Canada]	380,738	British Columbia mandatory health insurance database	4 years	NDVI yearly and seasonal; in 100 m circular buffer	Availability	Diabetes	Health register	OR: 0.90 [0.87, 0.92] IQR 1 increase in NDVI	sex, age, area-level household income, walkability, pollution	Good
Renzi et al. [63]	men and women; ≥35 years [Italy]	1,459,671	Rome Longitudinal Study	5.2 [mean] years	NDVI and LAI in a 300 m circular buffer	Availability	Diabetes	Medical records	β: −1.87 [−7.40, 3.99] Per unit increase in NDVI	SES, marital status, educational level, occupation, place of birth, sex	Good
Dalton et al. [64]	men and women; ≥40–80 years [United Kingdom]	25,633	European Prospective Investigation into Cancer [EPIC] Norfolk	11.3 [mean] years	Percent green space; in 800m circular buffer	Availability	Diabetes[T2]	Self-report of physician diagnosis or medication	HR: 0.81 [0.65, 0.99] Highest percent quintile green	sex, age, BMI, parental diabetes, SES	Good
Liao et al. [65]	pregnant women; 25–29 years mean age group [China]	6,883	Visitors of Wuhan’s Women and Children Medical and Healthcare Center	9 months or until development of gestational diabetes	NDVI for conception years; 300 m circular buffer	Availability	Diabetes [genstational]	Clinical samples	RR1: 0.66 [0.52, 0.84] Highest quintile NDVI	age, education years, BMI, passive smoking during pregnancy, parity, season	Good
Hobbs et al. [66]	men and women; ≥18–89 years [United Kingdom]	28,806	Yorkshire Health Study	3 years	Presence of park in a 2000 m circular buffer	Accessibility	Obesity	BMI, self-report	OR: 0.99 [0.98, 1.02] for answering “yes” to presence of park	age, sex, education, deprivation, population density	Fair
Persson et al. [67]	men and women, ≥35–65 years [Sweden]	5712	Stockholm Diabetes Prevention Program [SDPP]	8.9 [mean] years	NDVI; time-weighted in a 100 m, 250 m, 500 m circular buffer	Availability	Obesity	Objective measures of BMI	IRR for IQR increase in NDVI 500 m Females: 1.05 [0.88, 1.26] Males: 1.06 [0.89, 1.26]	age, alcohol consumption, tobacco use, psychological distress, shift work, aircraft noise, railway noise, distance to water	Good
Halonen et al. [68]	men and women; public sector employees; mean: 47.7 years [nonmovers] and among the movers 41.8 [Finland]	35,213	Finnish Public Sector study	8 years	Distance to green space; distance to blue space in meters, objectively measured	Accessibility	Obesity and overweight	Self-reported BMI	Green space OR: 1.50 [1.07, 2.11] Longest distance quintile Blue space OR: 1.15 [0.94, 1.39] Longest distance quintile	age, sex, education, chronic disease, neighbourhood socioeconomic disadvantage, BMI, smoking, heavy alcohol, physical inactivity	Poor
Lee et al. [69]	men and women; ≥19 years [48.6 years mean] [USA]	5435	Offspring and Generation Three Cohorts of the Framingham Heart Study	6.4 years	Percent green space within a census block	Availability	Obesity; Diabetes	Blood samples; medication; objectively-measured BMI	Diabetes: OR: 0.70 [0.41, 1.19] Highest percent quintile green Obesity: no results	age, gender, smoking status, education, cohort status, fasting plasma glucose, BMI	Fair
Astell-Burt and Feng [70]	men and women; ≥45 years [Australia]	53,196	45 and Up Study	6 years	Percent green space; tree canopy in a 1600 m road network buffer	Availability	Diabetes, hypertension and CVD	Self-report of physician diagnosis	Diabetes OR: 1.10 [0.65, 1.95] Highest percent quintile total green OR: 0.71 [0.56, 0.91] Highest percent quintile tree canopy Hypertension OR: 0.72 [0.64, 1.12] Highest percent quintile total green OR: 0.82 [0.71, 0.95] Highest percent quintile tree canopy CVD OR: 0.89 [0.59, 1.13] Highest percent quintile total green OR: 0.79 [0.63, 0.92] Highest quintile tree canopy	age, sex income, education, economic status, couple status	Poor
Paquet et al. [71]	men and women; ≥18 years [Australia]	4056	North West Adelaide Health Study [NWAHS]	3.5 [mean] years	NDVI in 1000 m road network buffer	Availability	Diabetes; hypertension; obesity; dyslipidaemia	Clinical samples	Per unit increase in NDVI Diabetes RR: 1.01 [0.90, 1.13] Hypertension RR: 0.97 [0.87, 1.07] Dyslipidaemia RR: 1.12 [1.00, 1.25] Obesity RR: 1.04 [0.92, 1.16]	age, gender, smoking status, education, cohort status, fasting plasma glucose, BMI	Good
de Keijzer et al. [72]	men and women; ≥35–55 years civil servants [United Kingdom]	10,308	Whitehall II	14.1 [median] years	NDVI and VCF, 500 m and 1000 m circular buffers and LSOA	Availability	Metabolic Syndrome	Clinical samples	IQR increase in NDVI 500m HR: 0.87 [0.77, 0.99] 1000 m HR: 0.90 [0.79, 1.01] LSOA HR: 0.91 [0.79, 1.03]	age, sex, ethnicity, individual socioeconomic status [education and employment grade], neighbourhood socioeconomic status [income and employment deprivation]	Good
Datzman et al. [73]	men and women; mean 49.33 years; [Germany]	1,918,449	AOK Plus [health insurance database]	4 years	NDVI; 115 images for 4 years; statistical area units	Availability	Cancer: colorectal; mouth and throat, prostate, breast; nonmelanoma skin	Health register	Per 10% increase in NDVI Colorectal: RR: 1.03 [0.98, 1.07] Mouth: RR: 0.89 [0.83, 0.96] Skin: RR: 0.84 [0.79, 0.90] Prostate: RR: 0.95 [0.90, 1.01] Breast: RR: 0.96 [0.92, 0.99]	age, sex, alcohol-related disorder, absolute number of physician contacts, proportion of short and long-term unemployment	Good
Conroy et al. [74]	women; ≥45–75 years; [African Americans, Japanese Americans, Latinos, Native Hawaiians, and White] [USA]	48,247	Multiethnic Cohort [MEC]	17 years	Presence of a park; based on number in a residential block group	Accessibility	Breast cancer [invasive]	Health register	HR: 1.03 [0.92, 1.15] No park in area	age, clustering effect of block group, ethnicity, risk factors, baseline BMI and adult weight change, neighbourhood SES, all neighbourhood obesogenic factors	Good
Haraldsdottir et al. [75]	women; mean: 53.9 years [Iceland]	10,049	Reykjavik Study	27.3 average	Coastal residence, self-reported	Availability	Breast cancer	Health registers	HR: 0.87 [0.72, 1.04] Coastal residence vs. city	age, birth cohort, education, physical activity, parity, height, BMI in midlife, age at menarche, age at first child	Good
Orioli et al. [76]	men and women; ≥30 years [Italy]	1,265,058	Rome Longitudinal Study	13 years	NDVI average for 2015 in 300 m and 1000 m circular buffer	Availability	Stroke	Health register	NDVI highest quintile 300 m HR: 0.95 [0.91, 0.98] 1000 m HR: 0.97 [0.93, 1.00]	age, sex, educational level, marital status, occupational status, place of birth, area-level SES	Good
Paul et al. [77]	men and women; ≥35–100 years; urban residents Ontario [Canada]	4,251,146	Ontario Population Health and Environment Cohort [ONPHEC]	13 years	NDVI annual values, 250 m circular buffer	Availability	Stroke	Health register	HR: 0.96 [95% CI: 0.95, 0.97] per IQR increase in NDVI	age, sex, SES, comorbidities, northern residence, population density, air pollution	Good
Yuchi et al. [78]	men and women; ≥45–84 years [Canada]	634,432 [parkinson disease]; 7232 [multimple sclerosis]	Medical Services Plan [MSP] Vancouver, mandatory health insurance database	4 years	NDVI; yearly average in 100 m circular buffer	Availability	Parkinson’s disease Multiple sclerosis	Health records	Per IQR increase in NDVI Parkinson’s Disease: OR: 0.97 [0.93, 1.01] Multiple Sclerosis: OR: 1.14 [1.00, 1.30]	Parkinson’s disease: age, sex, comorbidities, household income, education, ethnicity Multiple sclerosis: age, sex, comorbidities, household income, education and ethnicity, comorbidities, household income, education, ethnicity	Good
Picavet et al. [55]	men and women; ≥18 to 55 years [Netherlands]	4,917	Doetinchem Cohort Study	15 years	Percent green space in 125 m and 1000 m circular buffer	Availability	Obesity; Hypertension	All self-reported instruments	Per unit increase in percent green space 125 m Obesity: OR: 1.04 [1.01, 1.07] Hypertension: OR: 0.99 [0.97, 1.02] 1000 m Obesity: OR: 1.00 [0.96; 1.05] Hypertension: OR: 1.02 [0.98; 1.05]	age, sex, SES	Poor
Secondary Outcomes
de Keijzer et al. [79]	men and women; ≥35–55 civil servants [United Kingdom]	10,308	Whitehall II study	9 [median] years	NDVI and EVI; distance to blue space [any visible water]; distance to green or blue space in 500m and 1000 m circular buffer; distance in m	Availability Accessibility	Physical Functioning	Clinical measures	Walking speed [difference baseline and follow-up]: 500 m NDVI β: 0.02 [0.01, 0.04] per IQR increase 1000m NDVI β: 0.03 [0.01, 0.04] per IQR increase Blue space β: −0.01 [−0.02, 0.01] per IQR increase Grip strength [difference baseline and follow-up]: 500 m NDVI β: −0.01 [−0.03, 0.01] per IQR increase 1000 m NDVI β: −0.01 [−0.03, 0.01] per IQR increase Blue space β: −0.01 [−0.03, 0.01] per IQR increase	sex, ethnicity, marital status, height, alcohol use, intake of fruit and vegetables, smoking, rurality, education, employment grade, Index of Multiple Deprivation [IMD], income score and of the IMD, employment score	Fair
Yu et al. [80]	men and women; ≥65 years [Hong Kong]	4000	Mr and Ms Os Study	2 years	NDVI at baseline in a 300 m circular buffer	Availability	Frailty	Self-reported instrument	OR: 1.29 [1.04, 1.60] Highest quintile NDVI	age, sex, marital status, SES, current smoking status, alcohol intake, diet quality, baseline frailty status, number of diseases, cognitive function, physical activity, depression	Good
Zhu et al. [81]	men and women; ≥65 years [China]	34,342	Chinese Longitudinal Healthy Longevity Survey [CLHLS]	9 years	NDVI; annual averages for each year in 500 m buffer	Availability	Frailty	Self-reported instrument	OR: 1.02 [1.00, 1.04] Per unit increase in NDVI	age, sex, ethnicity, marital status, geographic region, urban or rural residence, education, occupation, financial support, social and leisure activity, smoking status, drinking status, physical activity	Good
Chong et al. [82]	men and women; ≥45 years with diabetes [T2] [Australia]	60,404	45 and Up Study and the follow-up Social, Economic and Environmental Factors [SEEF] Study	3.3 [mean] years	Percent green space in 500 m, 1000 m, and 2000 m road network buffer	Availability	Physical Activity	Self-reported instrument [MVPA: min/week]	Per highest percent quintile green 500 mMean: 0.61 [−0.26, 1.49] 1000 mMean: 0.94 [0.10, 1.79] 2000 mMean: 0.75 [0.03, 1.48]	age, sex, country of birth, education, disadvantage, physical functioning, BMI, psychological distress	Poor
Cleland et al. [83]	women parents; mean: 42.4 years;[Australia]	698	Children Living in Active Neighbourhoods [CLAN]	2 years	Amount of greenery and quality of parks, self-reported satisfaction	Availability Accessibility	Physical activity	Self-reported instrument [walking: for leisure and transport [min/ week]]	Amount of greeneryPersistently high vs. persistently low PA: RR: 1.80 [1.04, 3.13] Increased vs. persistently low PA: RR: 1.39 [0.90, 2.17] Quality of parks Persistently high vs. persistently low PA: RR: 1.73 [1.17, 2.57] Increased vs. persistently low PA: RR: 1.20 [0.89, 1.62]	age, marital status, number of children in the household, highest level of schooling	Poor
Coogan et al. [84]	women; ≥21–69 years; Black ethnicity [USA]	21,820	Black Women’s Health Study	2-6 years 98,280 person-years of follow-up.	Distance to park	Accessibility	Physical activity	Self-reported instrument [Walking for recreation and total walking: y/n]]	Recreation walking OR: 1.01 [0.89, 1.13] Shortest distance quintile Exercise walking OR: 1.01 [0.91, 1.12] Shortest distance quintile	age, region, BMI, smoking, alcohol, marital status, parity, caregiver status, residential moves, chronic conditions, history of cancer, moving residence, vacant housing, SES, crime	Poor
Dalton et al. [85]	men and women; mean age at baseline 62.2 [United Kingdom]	25,639	European Prospective Investigation into Cancer [EPIC] Norfolk	7.5 [mean] years	Percent green space at baseline for nonmovers; 800 m	Availability	Physical Activity	Self-reported instrument [Change in overall PA [hr/week]]	β: 4.21 [1.60, 6.81] Highest percent quintile green	age, sex, marital status, waist to hip ratio, BMI, morbidity, urban/ rural location	Fair
Faerstein et al. [86]	men and women; ≥18 years; civil servants [Brazil]	1731	Pro-Saude study	13 years	NDVI [800 m circular buffer]; presence of trees [visual inspection]; proximity to waterfronts;	Availability Accessibility	Physical activity	Self-reported instrument [nonwork PA: yes/no]	OR: 0.85 [0.44, 1.65] Highest quintile NDVI OR: 1.22 [0.62, 2.40] Highest percent quintile of trees OR: 2.46 [1.22, 4.93] Longest distance to waterfronts	sex, race, education, income, neighbourhood contextual variables	Poor
Hogendorf et al. [87]	men and women; mean: 53 years; [Netherlands]	4758	Gezondheid en Levens Omstandigheden Bevolking Eindhoven en omstreken [GLOBE]	10 years	Area of green space within a 1000 m circular buffer; Distance to green space	Availability Accessibility	Physical activity	Self-reported instrument [total walking and cycling: min/week]	Total walking and cycling Per ha increase in area of green β: 0.82 [−178.84, 180.48] Distance per 100m increase in green β: −22.36 [−46.19, 1.48]	marital status, income, employment, smoking, self-rated health	Poor
Josey and Moore. [88]	men and women; ≥25years; urban residents [Canada]	2707	Montreal Neighborhood Networks and Healthy Aging Panel [MoNNET-HA]	5 years	Distance to parks and green spaces	Accessibility	Physical Activity	Self-reported instrument [physical inactivity: y/n]	OR: 0.99 [0.99, 1.00] Per mile increase in distance	sex, age, self-reported health status, SES, household language, marriage status, residential duration, wave	Poor
Lin et al. [89]	men and women; ≥65–98 years [Hong Kong]	4000	OS and Ms. OS Study	7.8 [mean] years	NDVI in 300 m circular buffer	Availability	Physical activity	Self-reported instrument [Total PA score]	No relevant results	age, sex, marital status, education level, alcohol consumption, smoking, living alone, self-rated health, chronic conditions, functional impairment	Fair
Michael et al. [90]	men; ≥65 years [USA]	513	Neighborhoods and Physical Activity in Elderly Men	3.6 [mean] years	Distance to park	Accessibility	Physical activity	Self-reported instrument [walking: min/day]	RR for presence of park Low SES: 0.89 [0.70, 1.13] High SES: 1.34 [1.16, 1.55]	age, race education, occupation, marital status, self-reported health, BMI, smoking, drinking, chronic conditions	Fair
Sugiyama et al. [91]	men and women; mean: 54.4 years [Australia]	4802	AusDiab study	7 years	Park or nature reserve in the neighbourhood, self-reported	Accessibility	Physical Activity	Self-reported instrument [meeting PA guidelines: y/n]	OR: 0.96 [0.80, 1.15] for having a park in neighbourhood	age, sex, education, work status change, child change, mobility, BMI	Poor
Yang et al. [92]	men and women; ≥40–79 years [United Kingdom]	25,633	European Prospective Investigation into Cancer [EPIC] Norfolk	7 years	Presence of park or green space in 800 m circular buffer	Accessibility	Physical activity	Self-reported instrument [active commuting: y/n]	Park [yes]: OR: 1.30 [0.96, 1.74] Green space [yes]: OR: 1.12 [0.83, 1.53]	No adjustment	Poor
Meyer et al. [93]	men and women; ≥18–30 years; black and white [USA]	5115	Coronary Artery Risk Development in Young Adults [CARDIA]	13 years	Number of parks within a 3000 m circular buffer	Accessibility	Physical activity; Diet Quality	Self-reported validated instruments [PA: frequency walking, biking, running/ week]	No relevant results	N/A	Poor
Picavet et al. [55]	men and women; ≥18 to 55 years [Netherlands]	4917	Doetinchem Cohort Study	15 years	Percent green space in 125 m and 1000 m circular buffer	Availability	Physical activity; Quality of Life	All self-reported instruments [PA: meeting guidelines: y/n]	Per unit increase in NDVI 125 m Physical activity: OR: 1.02 [0.99; 1.04] Quality of Life: Mixed 1000 m Physical activity: OR: 1.01 [0.97; 1.05] Quality of Life: Mixed	age, sex, SES	Poor

¹ Abbreviations: BMI: Body Mass Index/CI: Confidence Intervals/HR: Hazard Ratio/IQR: Interquartile Range/MVPA: Moderate-to-vigorous physical activity/NDVI: Normalized Difference Vegetation Index/OR: Odds Ratio/PA: Physical activity/PM: Particulate matter/RR: Relative Risk/SES: Socioeconomic status/β: Beta coefficient; * Based on Newcastle–Ottawa Scale [NOS] for Cohort Studies.