Resource: Using greenspace in research on cognitive outcomes

Contributors: Alyssa Edling, Ingrid Jarvis, Michael Brauer, Sara D. Adar

This page serves as a guide for the selection and application of urban greenspace exposure measures in research on long-term cognitive outcomes, including reduced Alzheimer’s disease (AD) and AD-related dementia (ADRD) risk.

In Urban greenspace and dementia: Measures, mechanisms, and methodological guidance Jarvis et al. (2026) outline possible mechanistic pathways and recommend urban greenspace exposure measures tailored to each pathway. Hypothesized pathways include stress reduction, attention restoration, and improved mental health; increased physical activity; increased social cohesion; and reduced harmful environmental exposures. Careful selection of exposure metrics enables efficient investigation of hypothesized pathways, supports comparability between studies, and facilitates translation into actionable interventions.

The figure below, from Jarvis et al. (2026), outlines the connection between these pathways, greenspace exposure dimensions, data sources, and health outcomes. If you want to study one of these pathways, this page will provide methodological recommendations drawn from Jarvis et al. to guide you.

Greenspace Life Course Exposure Diagram

Figure 1 from Jarvis et al. (2026).1

How is greenspace measured?

  • Availability: The amount of greenspace in an area, such as the density of vegetation or percentage of vegetated land cover within a defined distance from participant’s residence.
  • Accessibility: The spatial proximity of greenspace to a location, such as distance or travel time to a park or presence or count of parks.
  • Visibility: The amount of greenspace visible from a person’s perspective from a location.

Table 1 from Jarvis et al. (2026).1

Type Source/dataset Spatial coverage/resolution Temporal coverage/resolution Exposure dimensions/measures Data access/pre-processed data Strengths Limitations
Vegetation Index Sentinel-2
(NDVI; EVI; LSU)		 Global
(10 m)		 2015 to present
(5 days)		 Availability: May calculate the minimum, maximum, or mean value within a buffer or administrative area (e.g. census tract, zip code) for a given year or multi-year period;
Change in vegetation index value over time		 Publicly available94 • Global coverage
• Higher-end moderate spatial resolution
• High temporal resolution and coverage
• Widely used measure (NDVI)		 • No indication of type, quality, or public accessibility (i.e., open for public use)
• Sensitive to time of year
• Type of greenspace captured may differ across regions
• Unit scale (NDVI and EVI) difficult to interpret and translate to policy
Vegetation index Landsat (NDVI; EVI; LSU) Global
(30 m)		 1984 to present
(16 days)		 Availability: May calculate the minimum, maximum, or mean value within a buffer or administrative area for a given year or multi-year period;
Change in vegetation index value over time		 Publicly available95, 96
Pre-processed:
• Canadian Urban Environmental Health Research Consortium (CANUE): NDVI (1984-2019) and LSU (1984-2016) summarized in buffers for Canada		 • Global coverage
• High temporal coverage and resolution
• Widely used measure (NDVI)		 • No indication of type, quality, or public accessibility (i.e., open for public use)
• Moderate spatial resolution
• Sensitive to time of year
• Type of greenspace captured may differ across regions
• Unit scale (NDVI and EVI) difficult to interpret and translate to policy
Vegetation Index MODIS (NDVI; EVI; LSU) Global
(250 m)		 2000 to present
(Daily)		 Availability: May calculate the minimum, maximum, or mean value within a buffer or administrative area for a given year or multi-year period;
Change in vegetation index value over time		 Publicly available96-98
Pre-processed:
• Gateway to Global Aging Data: NDVI (2000-2022) summarized in buffers for eight countries in the Health and Retirement Study International Network of Studies
• CANUE: NDVI (2000-2023) summarized in buffers for Canada		 • Global coverage
• High temporal coverage and resolution
• Widely used measure (NDVI)		 • No indication of type, quality, or public accessibility
• Low spatial resolution
• Sensitive to time of year
• Type of greenspace captured may differ across regions
• Unit scale difficult to interpret and translate to policy
Tree cover Multi-Resolution Land Characteristics Consortium
(NLCD Tree Canopy Cover)		 United States
(30 m)		 1985 to 2023
(Annual)		 Availability: May calculate the percent or area of tree canopy within a buffer or administrative area for a given year or multi-year period;
Percent change in tree canopy over time		 Publicly available99 • Indication of tree exposure (compared to ‘greenness’)
• Nationwide coverage
• High temporal coverage
• Unit scale easy to interpret and translate to policy		 • No indication of quality or public accessibility
• Moderate spatial resolution
• Moderate temporal resolution
• No indication of species, seasonal foliage, or vegetation structure
Tree cover University of Maryland
(VCF)		 Global
(30 m)		 2000 to 2015
(Every 5 years: 2000, 2005, 2010, 2015)		 Availability: May calculate the percent or area of tree canopy within a buffer or administrative area for a given year or multi-year period;
Percent change in tree canopy over time		 Publicly available100 • Indication of tree exposure (compared to ‘greenness’)
• Global coverage
• High temporal coverage
• Unit scale easy to interpret and translate to policy		 • No indication of quality or public accessibility
• Moderate spatial resolution
• Low temporal resolution
• No indication of species, seasonal foliage, or vegetation structure
Tree cover University of Maryland
(VCF)		 Global
(250 m)		 2000 to present
(Annual)		 Availability: May calculate the percent or area of tree canopy within a buffer or administrative area for a given year or multi-year period;
Percent change in tree canopy over time		 Publicly available101 • Indication of tree exposure (compared to ‘greenness’)
• Global coverage
• High temporal coverage
• Unit scale easy to interpret and translate to policy		 • No indication of quality or public accessibility
• Low spatial resolution
• Moderate temporal resolution
• No indication of species, seasonal foliage, or vegetation structure
Tree cover Meta and WRI
(Global Canopy Height Map)		 Global
(1 m)		 Data spanning 2018 to 2020
(Annual)		 Availability: May calculate the percent or area of tree canopy within a buffer or administrative area for a given year Publicly available96, 102
Pre-processed:
• CANUE: Tree canopy cover (2021) summarized in buffers for 129 cities in Canada		 • Indication of tree exposure (compared to ‘greenness’)
• Global coverage
• High spatial resolution
• Unit scale easy to interpret and translate to policy		 • No indication of quality or public accessibility
• Low temporal coverage and resolution
• No indication of species, seasonal foliage, or vegetation structure
Tree cover ETH Zurich
(Global Canopy Height)		 Global
(10 m)		 2020 Availability: May calculate the percent or area of tree canopy within a buffer or administrative area for a given year Publicly available103 • Indication of tree exposure (compared to ‘greenness’)
• Global coverage
• Higher-end moderate spatial resolution
• Unit scale easy to interpret and translate to policy		 • No indication of quality or public accessibility
• Low temporal coverage and resolution
• No indication of species, seasonal foliage, or vegetation structure
Tree cover Global Land Analysis and Discovery
(Global Forest Change)		 Global
(30 m)		 2000 to 2024
(Updated annually)		 Availability: May calculate the percent of tree canopy cover within a buffer or administrative area for the year 2000;
Loss or gain in tree canopy cover from 2000 onwards		 Publicly available104 • Indication of tree exposure (compared to ‘greenness’)
• Estimate of forest loss and gain
• Global coverage
• High temporal coverage and resolution for forest change
• Unit scale easy to interpret and translate to policy		 • No indication of quality or public accessibility
• Moderate spatial resolution
• Low temporal coverage and resolution for forest extent
• No indication of species, seasonal foliage, or vegetation structure
Land cover Multi-Resolution Land Characteristics Consortium
(NLCD)		 United States
(30 m)		 1985 to 2024
(Annual)		 Availability: May calculate the percent or area of aggregated green land cover and specific land cover types within a buffer or administrative area for a given year;
Change in land cover over time		 Publicly available99, 105
Pre-processed:
• National Neighborhood Data Archive (NaNDA): NLCD (1985-2023) summarized in US census tracts and ZIP code tabulation areas (ZCTA)		 • Distinguish different vegetated and other land cover types (n = 16)
• Nationwide coverage
• High temporal coverage
• Unit scale easy to interpret and translate to policy		 • No indication of quality or public accessibility
• Moderate spatial resolution
• Moderate temporal resolution
• Classes may not align with other regional land cover datasets
• Developed land cover classes include mixture of man-made structures and vegetation
• Vegetated land cover classes do not indicate species or vegetation structure
Land cover US EPA EnviroAtlas
(Land cover)		 Select US cities (n = 30)
(1 m)		 Data spanning 2008 to 2013
(Available years vary across community layers)		 Availability: May calculate the percent or area of aggregated green land cover and specific land cover types within a buffer or administrative area for a given year Publicly available106
Pre-processed:
• US EPA: Greenspace, tree, agriculture, and wetland cover (2008-2013) summarized in US census block groups for 30 urban community areas		 • Distinguish different vegetated and other land cover types (n = 8)
• High spatial resolution
• Unit scale easy to interpret and translate to policy		 • No indication of quality or public accessibility
• Only select US communities
• Low temporal coverage and resolution
• Included classes vary across community layers
• Classes may not align with other regional land cover datasets
• Vegetated land cover classes do not indicate species, seasonal foliage, or vegetation structure
Land cover Commission for Environmental Cooperation
(North American Environmental Atlas)		 North America
(30 m and 250 m)		 2005 to 2020
(Every 5 years (i.e., 2005, 2010, 2015, 2020)		 Availability: May calculate the percent or area of aggregated green land cover and specific land cover types within a buffer or administrative area for a given year;
Change in land cover over time		 Publicly available107 • Distinguish different vegetated and other land cover types (n = 19)
• North American (US, Canada, Mexico) coverage
• High temporal coverage
• Unit scale easy to interpret and translate to policy		 • No indication of quality or public accessibility
• Moderate (30 m) and low (250 m) spatial resolution
• Low temporal resolution
• Classes may not align with other regional land cover datasets
• Vegetated land cover classes do not indicate species, seasonal foliage, or vegetation structure
Land cover European Environment Agency
(CORINE)		 Europe
(100 m)		 1990 to 2018
(Every 6 years (i.e., 1990, 2000, 2006, 2012, 2018))		 Availability: May calculate the percent or area of green urban areas, aggregated green land cover, and specific land cover types within a buffer or administrative area for a given year;
Change in land cover over time
Accessibility: May calculate distance or travel time to green urban area from a location;
Presence or count of green urban areas within a distance		 Publicly available108 • Distinguish different vegetated and other land cover types (n = 44)
• European coverage
• High temporal coverage
• Unit scale easy to interpret and translate to policy		 • No indication of quality or public accessibility
• Low spatial resolution
• Low temporal resolution
• Classes may not align with other regional land cover datasets
• Urban fabric land cover classes include mixture of man-made structures and vegetation
• Vegetated land cover classes do not indicate species or vegetation structure
Land use/ land cover EU
(Urban Atlas)		 Select European cities (n = 788 Functional Urban Areas)
(Minimum mapping width 10 m)		 2006 to 2018
(Every 6 years (i.e., 2006, 2012, 2018))		 Availability: May calculate the percent or area of green urban areas, aggregated green land cover, and specific land cover types within a buffer or administrative area for a given year;
Change in land cover over time
Accessibility: May calculate distance or travel time to green urban area from a location;
Presence or count of green urban areas within a distance		 Publicly available109 • Distinguish different vegetated and other land cover types (n = 27)
• Higher-end moderate spatial resolution
• High temporal coverage
• Unit scale easy to interpret and translate to policy		 • No indication of quality or public accessibility
• Only select European cities
• Low temporal resolution
• Classes may not align with other regional land cover datasets
• Urban fabric land cover classes include mixture of man-made structures and vegetation
• Vegetated land cover classes do not indicate species or vegetation structure
Land cover ESA
(WorldCover)		 Global
(10 m)		 2020 (V1) and 2021 (V2) Availability: May calculate the percent or area of aggregated green land cover and specific land cover types within a buffer or administrative area for a given year Publicly available110, 111 • Distinguish different vegetated and other land cover types (n = 10)
• Global coverage
• Higher-end moderate spatial resolution
• Unit scale easy to interpret and translate to policy		 • No indication of quality or public accessibility
• Low temporal coverage and resolution
• Vegetated land cover classes do not indicate species, seasonal foliage, or vegetation structure
• Versions use different algorithm so cannot reliably estimate change
Land cover Esri
(Land cover)		 Global
(10 m)		 2017 to 2024
(Annual)		 Availability: May calculate the percent or area of aggregated green land cover and specific land cover types within a buffer or administrative area for a given year;
Change in land cover over time		 Publicly available112 • Distinguish different vegetated and other land cover types (n = 9)
• Global coverage
• Higher-end moderate spatial resolution
• Unit scale easy to interpret and translate to policy		 • No indication of quality or public accessibility
• Moderate temporal coverage and resolution
• Built area class includes a mixture of man-made structures and vegetation
• Vegetated land cover classes do not indicate species, seasonal foliage, or vegetation structure
Land cover Google and WRI
(Dynamic World)		 Global
(10 m)		 2015 to present
(every 2-5 days)		 Availability: May calculate the percent or area of aggregated green land cover and specific land cover types within a buffer or administrative area for a given year;
Change in land cover over time		 Publicly available113 • Distinguish different vegetated and other land cover types (n = 9)
• Global coverage
• Higher-end moderate spatial resolution
• High temporal coverage and resolution
• Unit scale easy to interpret and translate to policy		 • No indication of quality or public accessibility
• Built area class includes a mixture of man-made structures and vegetation
• Vegetated land cover classes do not indicate species, seasonal foliage, or vegetation structure
Land use Browning et al. and United States Geological Survey (USGS) data
(PAD-US-AR)		 United States
(Polygons)		 2020 (last update) Availability: May calculate the percent or area of park land within a buffer or administrative area;
Accessibility: May calculate distance or travel time to a park from a location;
Presence or count of parks within a distance		 Publicly available114 • Identifies publicly accessible parks intended for recreation
• Nationwide coverage
• Unit scale easy to interpret and translate to policy		 • No indication of quality (e.g., amenities, cleanliness, safety)
• Low temporal coverage and resolution
Land use Esri
(USA Parks)		 United States
(Polygons)		 2025 (last update) Availability: May calculate the percent or area of park land within a buffer or administrative area;
Accessibility: May calculate distance or travel time to a park from a location;
Presence or count of parks within a distance		 Publicly available115 • Identifies national and state parks and forests, and county, regional, and local parks
• Nationwide coverage
• Unit scale easy to interpret and translate to policy		 • No indication of public accessibility or quality (e.g., amenities, cleanliness, safety)
• Low temporal coverage and resolution
Land use Trust for Public Land
(ParkServe)		 Urban areas in the US, including Puerto Rico
(Polygons)		 2025 (last update)
(Updated annually)		 Availability: May calculate the percent or area of park land within a buffer or administrative area;
Accessibility: May calculate distance or travel time to a park from a location;
Presence or count of parks within a distance		 Publicly available116, 117
Pre-processed:
• Trust for Public Land: 10-minute walk service area of parks (last update) for all US urban areas
• NaNDA: ParkServe (2018, 2022) total number, area, and proportion of parks within each US census tract or ZCTA		 • Identifies publicly accessible local parks
• Includes basic amenity (i.e., trails, playgrounds) and vegetation (i.e., percent tree canopy cover) data for parks
• Nationwide coverage for urban areas
• Unit scale easy to interpret and translate to policy		 • No indication of quality (e.g., cleanliness, safety)
• Low temporal coverage and resolution
Land use Open Street Map Foundation
(OSM)		 Global
(Polygons)		 Continuous updates Availability: May calculate the percent or area of park land within a buffer or administrative area;
Accessibility: May calculate distance or travel time to a park from a location;
Presence or count of parks within a distance		 Publicly available118 • May identify possible parks through data features tagged by users
• May identify park amenities (e.g., path, toilet, picnic table) through data features tagged by users
• Unit scale easy to interpret and translate to policy		 • No indication of quality (e.g., cleanliness, safety)
• Uneven spatial coverage and completeness
• Uneven temporal coverage and resolution
• Accuracy and consistency of tagged features varies globally
Visibility Google
(Street View Imagery)		 Global
(Resolution varies based on camera system and ground sampling)		 2007 to present
(Available years and updates vary by region)		 Visibility: May calculate the percent or area of aggregated green or individual vegetation features within an image Restricted119 • Identify visible greenspace features at ground level
• Unit scale easy to interpret and translate to policy		 • Uneven spatial coverage and resolution
• Uneven temporal coverage and resolution
• Require high computational skills and power to process images
• Restrictions on terms of use

Urban greenspace exposure measurement recommendations

The recommendations below align urban greenspace exposure measures with distinct hypothesized pathways linking greenspace to AD/ADRD outcomes. NDVI is presented first as a practical baseline measure, followed by the recommended greenspace exposure measures ordered according to their hypothesized relevance for capturing each pathway.

  1. Stress reduction, attention restoration, and improved mental health outcomes
  2. Increased physical activity
  3. Increased social cohesion
  4. Reduced harmful environmental exposures

Pathway: Stress reduction, attention restoration, and improved mental health outcomes

Research has connected greenspace to short-term physiological changes and improvements in attention, mental fatigue and mood.2 These short term changes may then be connected to moderate-term outcomes like improved mental health, which is a known modifiable risk factor for AD/ADRD.3 Stress Reduction Theory (SRT) and Attention Restoration Theory (ART) are foundational theories emphasizing the role of greenspace in recovery from stress and attentional fatigue.4,5

Exposure: Residential greenspace availability

NDVI is not a pathway-specific measure but is a common measure that may be used as a starting point across all pathways.

  • Rationale: A small circular buffer captures greenspace availability and visibility surrounding the participant’s residence.
  • Measure specifications:
    • Sensitivity: 50-meter circular buffer
    • Recommended data source: Landsat satellite imagery (30 m)
    • Alternate data sources: Sentinel (10 m)

  • Rationale: A large network buffer captures greenspace exposure on travel routes that participants may intentionally use within the neighborhood.
  • Measure specifications:
    • Sensitivity: 800-meter network buffer
    • Recommended data source: Landsat satellite imagery (30 m)
    • Alternate data sources: Sentinel (10 m), MODIS (250 m)

  • NDVI is a globally comparable measure and has been found to correspond to expert ratings of ‘greenness.’6
  • Landsat imagery is publicly available with wide spatial and temporal coverage.
  • NDVI has been widely applied in observational studies, facilitating consistency and enabling meta-analysis.
  • NDVI does not provide info on type, quality, or public accessibility of greenspace.
  • Differences of greenspace measures across regions can complicate cross-country analysis and comparisons.
  • NDVI may differ from natural spaces and may be sensitive to time of year.
  • NDVI’s unit scale (-1 to 1) is difficult to interpret for policy translation.

Exposure: Residential tree canopy cover availability

Trees may enhance aesthetic appeal and have been more consistently linked with stress reduction and positive mental health outcomes than other greenspace measures.7-9

  • Rationale: Visibility would ideally be assessed as it may be a relevant form of exposure for restorative benefits, but such data are not widely available. In absence of a direct measure of visibility, tree canopy cover within a small circular buffer zone may serve as a proxy, assuming nearby trees are partially visible.
  • Measure specifications:
    • Sensitivity measure: 50-meter circular buffer
    • Data sources: NLCD TCC, EPA EnviroAtlas, U of Maryland's Vegetation Continuous Field (VCF), Meta and World Resources Institute Global Canopy Height Map, ETH Zurich Global Canopy Height, Global Land Analysis and Discovery's Global Forest Change, NLCD, North American Environmental Atlas, CORINE, Urban Atlas, World Cover, Esri Land Cover, Dynamic World

  • Rationale: A large network buffer captures tree canopy cover along travel routes that participants may intentionally use within the neighborhood.
  • Measure specifications:
    • Sensitivity measure: 800-meter buffer
    • Data Sources: NLCD TCC, EPA EnviroAtlas, U of Maryland's Vegetation Continuous Field (VCF), Meta and World Resources Institute Global Canopy Height Map, ETH Zurich Global Canopy Height, Global Land Analysis and Discovery's Global Forest Change, NLCD, North American Environmental Atlas, CORINE, Urban Atlas, World Cover, Esri Land Cover, Dynamic World

  • Provides an indication of vegetation type rather than just greenness
  • Tree canopy cover research translates to actionable urban planning and management insights, like tree planting initiatives.
  • Tree canopy cover does not provide information on the quality or accessibility of greenspace or fully represent the human-view visibility perspective.
  • Land cover data classifications can lack detail and introduce possible measurement errors, and regional data sets may lack comparability.
  • Tree canopy cover overlooks possible restorative properties of other vegetation, like flowers, and other natural environments, like deserts.10

Exposure: Park Accessibility

Parks may also be a helpful exposure as they encourage engagement in physical activity that supports mental well-being.

  • Rationale: Network distance or travel time, ideally measured relative to the park’s entry point, may serve as proxies for park accessibility. Since research suggests activity levels may depend on park size, it may be useful to set a minimum size threshold.11,12 In line with the WHO recommendations, 0.5 hectares is recommended as the primary park size threshold.13,14
  • Measure specifications:
    • Sensitivity measure: 1 hectare park size
    • Data sources: Parks and Protected Areas Database, Esri USA Parks, Trust for Public Land’s ParkServe

  • Rationale: When measuring park presence or total park number, use a small distance as parks often need to be relatively close to support familiarity and regular use. The 300-meter distance aligns with the WHO recommendation based on a 5-minute walk.13,14 Again, 0.5 hectares is the recommended primary park threshold.
  • Measure specifications:
    • Sensitivity measure: 800-meter buffer (i.e.,10-minute walk), 1 hectare park size
    • Data sources: Parks and Protected Areas Database, Esri USA Parks, Trust for Public Land’s ParkServe

  • Park accessibility enables actionable urban planning and management insights.
  • Park proximity does not guarantee use.
  • The size, quality, safety, and amenities of parks is not captured and may determine visitation more than proximity.11,15
  • Park data may not indicate public accessibility to freely use the space.
  • Park use and preferences differ among populations and may be determined by factors other than physical activity to support mental well-being, such as dog ownership.16–18
  • Means other than walking may be used to reach parks, which may change the appropriate travel distance.
  • Distance may not reflect ease of travel (e.g. hilly terrain, uneven surfaces, barriers).

Key Literature
  • Astell-Burt and Feng (2023): This study measured tree canopy cover within 1.6km of the home, a large buffer to cover standard walking distance, and used hospital and mortality records to assess dementia cases. Absence of psychological distress mediated 5.7% of the positive association between tree canopy cover and dementia, the largest impact of any examined pathway
  • Lega et al. (2021): This study explored the relationship between greenness, as measured by NVDI, self-perceived stress, and cognitive test scores. Greenness was associated with better short-term and overall memory, and reduced stress was a significant partial mediator.
  • Jimenez et al. (2022): This study also used NVDI to associate greenspace and cognitive testing scores in mid-life. Depressive symptoms partially mediated the positive association between greenspace and cognition.

Back to the top or recommendation list

Pathway: Increased physical activity

Greenspace may support and encourage healthy behaviors, such as physical activity, which is a known modifiable risk factor for dementia.3,22,23 Physical activity supported by greenspace can include walking through parks, gardening, or walking on tree-shaded streets.

Exposure: Residential greenspace availability

NDVI is not a pathway-specific measure but is a common measure that may be used as a starting point across all pathways.

  • Rationale: A large network buffer captures greenspace exposure on travel routes that participants may intentionally use for physical activity within the neighborhood.
  • Measure specifications:
    • Sensitivity: 800-meter network buffer
    • Recommended data source: Landsat satellite imagery (30 m)
    • Alternate data sources: Sentinel (10 m), MODIS (250 m)

  • NDVI is a globally comparable measure and has been found to correspond to expert ratings of ‘greenness.’6
  • Landsat imagery is publicly available with wide spatial and temporal coverage.
  • NDVI has been widely applied in observational studies, facilitating consistency and enabling meta-analysis.
  • NDVI does not provide info on type, quality, or public accessibility of greenspace.
  • Differences of greenspace measures across regions can complicate cross-country analysis and comparisons.
  • NDVI may differ from natural spaces and may be sensitive to time of year.
  • NDVI’s unit scale (-1 to 1) is difficult to interpret for policy translation.

Exposure: Park accessibility

Prior studies have supported the positive relationship between park access and physical activity. 24,25

  • Rationale: Network distance or travel time, ideally measured relative to the park’s entry point, may serve as proxies for park accessibility. Since research suggests activity levels may depend on park size, it may be useful to set a minimum size threshold.11,12 In line with the WHO recommendations, 0.5 hectares is recommended as the primary park size threshold.13,14
  • Measure specifications:
    • Sensitivity measure: 1 hectare park size
    • Data sources: Parks and Protected Areas Database, Esri USA Parks, Trust for Public Land’s ParkServe

  • Rationale: When measuring park presence or total park number, use a small distance as parks often need to be relatively close to support familiarity and regular use. The 300-meter distance aligns with the WHO recommendation based on a 5-minute walk. 13,14 Again, 0.5 hectares is the recommended primary park threshold.
  • Measure specifications:
    • Sensitivity measure: 800-meter buffer (i.e.,10-minute walk), 1 hectare park size
    • Data sources: Parks and Protected Areas Database, Esri USA Parks, Trust for Public Land’s ParkServe

  • Park accessibility enables actionable urban planning and management insights.
  • Park proximity does not guarantee use.
  • The size, quality, safety, and amenities of parks is not captured and may determine visitation more than proximity.11,15
  • Park data may not indicate public accessibility to freely use the space.
  • Park use and preferences differ among populations and may be determined by factors other than physical activity, such as dog ownership.16–18
  • Means other than walking may be used to reach parks, which may change the appropriate travel distance.
  • Distance may not reflect ease of travel (e.g. hilly terrain, uneven surfaces, barriers).

Exposure: Residential tree canopy cover availability

Tree canopy cover outside of parks may also encourage physical activity, for example walking along tree-shaded streets.

  • Rationale: A large network buffer captures tree canopy cover along travel routes that participants may intentionally use for physical activity within the neighborhood.
  • Measure specifications:
    • Sensitivity measure: 800-meter buffer
    • Data Sources: NLCD TCC, EPA EnviroAtlas, U of Maryland's Vegetation Continuous Field (VCF), Meta and World Resources Institute Global Canopy Height Map, ETH Zurich Global Canopy Height, Global Land Analysis and Discovery's Global Forest Change, NLCD, North American Environmental Atlas, CORINE, Urban Atlas, World Cover, Esri Land Cover, Dynamic World

  • Provides an indication of vegetation type rather than just greenness.
  • Tree canopy cover research translates to actionable urban planning and management insights, like tree planting initiatives.
  • Tree canopy cover does not provide information on the quality or accessibility of greenspace or fully represent the human-view visibility perspective.
  • Land cover data classifications can lack detail and introduce possible measurement errors, and regional data sets may lack comparability.
  • Tree canopy cover overlooks possible restorative properties of other vegetation, like flowers, and other natural environments, like deserts.10

Key Literature

  • Sylvers et al. 2022 This study used both subjective perception and objective measures of park availability, cognitive test scores, and survey responses on physical activity to examine the role of neighborhood design in healthy aging. Moderate physical activity mediated the increase in cognitive score associated with the subjective perception of adequate parks, but vigorous activity and walking behavior did not.

  • Finlay et al. 2021 This blended qualitative and quantitative study links neighborhood features that may encourage physical activity with cognitive test outcomes. Many respondents described parks as beneficial places to exercise, and those living in areas with more parks scored higher on cognitive function tests.

  • de Keijzer et al. 2018 This longitudinal study looked at associations between residential greenness, measured by NDVI, cognitive test score changes, and survey responses on physical activity. A relationship was found between greenness and slower cognitive decline, but physical activity was not a mediator of this relationship.

Back to the top or recommendation list

Pathway: Increased social cohesion

Greenspaces provide space for direct and ambient social engagement, facilitating social cohesion. Social interaction is a known modifiable risk factor for dementia.3 Research examining social cohesion as a possible mediator for the association between greenspace and cognitive outcomes is currently limited.

Exposure: Residential greenspace availability

NDVI is not a pathway-specific measure but is a common measure that may be used as a starting point across all pathways.

  • Rationale: A large network buffer captures greenspace exposure on travel routes that participants may intentionally use for social interactions within the neighborhood.
  • Measure specifications:
    • Sensitivity: 800-meter network buffer
    • Recommended data source: Landsat satellite imagery (30 m)
    • Alternate data sources: Sentinel (10 m), MODIS (250 m)

  • NDVI is a globally comparable measure and has been found to correspond to expert ratings of ‘greenness.’6
  • Landsat imagery is publicly available with wide spatial and temporal coverage.
  • NDVI has been widely applied in observational studies, facilitating consistency and enabling meta-analysis.
  • NDVI does not provide info on type, quality, or public accessibility of greenspace.
  • Differences of greenspace measures across regions can complicate cross-country analysis and comparisons.
  • NDVI may differ from natural spaces and may be sensitive to time of year.
  • NDVI’s unit scale (-1 to 1) is difficult to interpret for policy translation.

Exposure: Park accessibility

Parks, in particular, may encourage engagement with friends, family, and neighbors and proximity to parks reflects ease of access. Prior research suggests that familiarity with parks increases the likelihood of interacting with others in them.28–30

  • Rationale: Network distance or travel time, ideally measured relative to the park’s entry point, may serve as proxies for park accessibility. Larger parks do not necessarily lead to more interaction, and small pocket parks may even be more frequently visited for planned and unplanned social activities, so no park-size limit is recommended.31
  • Measure specifications:
    • Data sources: Parks and Protected Areas Database, Esri USA Parks, Trust for Public Land’s ParkServe

  • Rationale: When measuring park presence or total park number, use a small distance as parks often need to be relatively close to support familiarity and regular use. The 300-meter distance aligns with the WHO recommendation based on a 5-minute walk.13,14
  • Measure specifications:
    • Sensitivity measure: 800-meter buffer (i.e.,10-minute walk)
    • Data sources: Parks and Protected Areas Database, Esri USA Parks, Trust for Public Land’s ParkServe

  • Park accessibility enables actionable urban planning and management insights
  • Park proximity does not guarantee use.
  • The size, quality, safety, and amenities of parks is not captured and may determine visitation more than proximity.11,15
  • Park data may not indicate public accessibility to freely use the space.
  • Certain amenities such as benches and tables or programming for social gatherings, as well as perceived safety and crime, may be particularly important for facilitating social interaction.32
  • Park use and preferences differ among populations and may be determined by factors other than social activity, such as dog ownership.16–18
  • Means other than walking may be used to reach parks, which may change the appropriate travel distance.
  • Distance may not reflect ease of travel (e.g., hilly terrain, uneven surfaces, barriers).

Exposure: Residential tree canopy cover availability

You may also use tree canopy cover as a measure for social connection. Prior research suggests social support may mediate part of the relationship between tree canopy cover and dementia risk.19

  • Rationale: A large network buffer captures tree canopy cover along travel routes that participant’s may intentionally use for social interaction within the neighborhood.
  • Measure specifications:
    • Sensitivity measure: 800-meter network buffer
    • Data Sources: NLCD TCC, EPA EnviroAtlas, U of Maryland's Vegetation Continuous Field (VCF), Meta and World Resources Institute Global Canopy Height Map, ETH Zurich Global Canopy Height, Global Land Analysis and Discovery's Global Forest Change, NLCD, North American Environmental Atlas, CORINE, Urban Atlas, World Cover, Esri Land Cover, Dynamic World

  • Provides an indication of vegetation type rather than just greenness.
  • Tree canopy cover research translates to actionable urban planning and management insights, like tree planting initiatives.
  • Tree canopy cover does not provide information on the quality or accessibility of greenspace or fully represent the human-view visibility perspective.
  • Land cover data classifications can lack detail and introduce possible measurement errors, and regional data sets may lack comparability.
  • Tree canopy cover overlooks possible restorative properties of other vegetation, like flowers, and other natural environments, like deserts.10

Key Literature

  • Astell-Burt et al. 2023 This study measured tree canopy cover within 1.6km of the home, hospital and mortality records to assess dementia cases, and surveys to assess social support and loneliness. Social support was a partial mediator of this relationship; however, social loneliness was not a mediator.

  • Zijlema et al. 2017 This study explored the relationships between distance to natural outdoor environments (NOE), surrounding greenness measured by NVDI, cognitive function test results, and survey responses to assess social cohesion. It found a relationship between the distance to NOE and cognitive test completion time, but social interaction did not mediate that relationship. No relationship was found between greenness measured by NVDI and cognitive function

Back to the top or recommendation list

Pathway: Reduced levels of harmful environmental exposures

Harmful environmental exposures like air pollution, noise, and heat may be reduced by greenspace.3,34,35 Spaces with high greenspace have fewer source emissions (e.g. traffic, industrial sites) and greenspace may even actively mitigate harmful exposures.36–38 Vegetation also reduces heat through shading, evapotranspiration, and the absence of paved surfaces.39–41

Exposure: Residential greenspace availability

NDVI is not a pathway-specific measure but is a common measure that may be used as a starting point across all pathways. For reduction of harmful environmental exposures, NVDI is also specifically recommended as multiple facets of greenspace may mitigate environmental harm.

  • Rationale: To examine localized neighborhood-level reduction and/or mitigation of harmful exposures, use a large circular buffer.
  • Measure specifications:
    • Sensitivity measure: 1,000-meter circular buffer
    • Recommended data source: Landsat satellite imagery (30 m)
    • Alternate data sources: Sentinel (10 m), MODIS (250 m)

  • Rationale: For localized heat reduction and noise buffering, use a smaller circular buffer.
  • Measure specifications:
    • Sensitivity measure: 50-meter circular buffer
    • Recommended data source: Landsat satellite imagery (30 m)
    • Alternate data sources: Sentinel (10 m)

Exposure: Residential tree canopy cover availability

Tree canopy cover may also be used for harmful exposures as taller vegetation offers more shade and row trees may function as a noise barrier.37

  • Rationale: To examine neighborhood-level reduction and/or mitigation of harmful exposures, use a large circular buffer.
  • Measure specifications:
    • Sensitivity measure: 1,000-meter buffer
    • Data sources: NLCD TCC, EPA EnviroAtlas, U of Maryland's Vegetation Continuous Field (VCF), Meta and World Resources Institute Global Canopy Height Map, ETH Zurich Global Canopy Height, Global Land Analysis and Discovery's Global Forest Change, NLCD, North American Environmental Atlas, CORINE, Urban Atlas, World Cover, Esri Land Cover, Dynamic World

  • Rationale: Use a small buffer to examine tree canopy cover’s localized impacts on residential noise reduction and cooling.
  • Measure specifications:
    • Sensitivity measure: 50-meter buffer
    • Data Sources: NLCD TCC, EPA EnviroAtlas, U of Maryland's Vegetation Continuous Field (VCF), Meta and World Resources Institute Global Canopy Height Map, ETH Zurich Global Canopy Height, Global Land Analysis and Discovery's Global Forest Change, NLCD, North American Environmental Atlas, CORINE, Urban Atlas, World Cover, Esri Land Cover, Dynamic World

  • Provides an indication of vegetation type rather than just greenness.
  • Tree canopy cover research translates to actionable urban planning and management insights, like tree planting initiatives.
  • Both NVDI and tree canopy cover lack information on species, seasonal variation, or structure, which can influence air pollution and cooling efficiency 36,41,42
  • Tree species have health risk pathways as allergens.43
  • Under some circumstances, greenspace may reduce dispersion, alter wind, or release volatile organic compounds, worsening pollution.44

Key Literature
  • Crous-Bou et al. (2020): This study evaluated the effect of urban environmental exposures on cognitive performance, using NVDI and land use type with cognitive scores and brain imaging. It found that increasing residential greenness was associated with greater cortical thickness, and associations were mediated by nitrogen oxides and particulate matter.
  • Besser et al. (2021): This study used park distance, proportion of parks, and cognitive test scores to assess the relationship between park access and cognitive change. It found a positive association between cognitive test score trajectory and neighborhood park availability. Air pollution was evaluated as a possible mediator, but there was no evidence of mediation.

Back to the top or recommendation list

References

  1. Jarvis I, Besser L, Finlay J, et al. Urban greenspace and dementia: Measures, mechanisms, and methodological guidance. Alzheimer’s & Dementia. 2026;22(6):e71557. doi:10.1002/alz.71557

  1. Song S, Tu R, Lu Y, Yin S, Lin H, Xiao Y. Restorative Effects from Green Exposure: A Systematic Review and Meta-Analysis of Randomized Control Trials. IJERPH. 2022;19(21):14506. doi:10.3390/ijerph192114506

  1. Livingston G, Huntley J, Liu KY, et al. Dementia prevention, intervention, and care: 2024 report of the Lancet standing Commission. The Lancet. 2024;404(10452):572-628. doi:10.1016/S0140-6736(24)01296-0

  1. Kaplan S. The restorative benefits of nature: Toward an integrative framework. Journal of Environmental Psychology. 1995;15(3):169-182. doi:10.1016/0272-4944(95)90001-2

  1. Ulrich RS, Simons RF, Losito BD, Fiorito E, Miles MA, Zelson M. Stress recovery during exposure to natural and urban environments. Journal of Environmental Psychology. 1991;11(3):201-230. doi:10.1016/S0272-4944(05)80184-7

  1. Rhew IC, Vander Stoep A, Kearney A, Smith NL, Dunbar MD. Validation of the Normalized Difference Vegetation Index as a Measure of Neighborhood Greenness. Annals of Epidemiology. 2011;21(12):946-952. doi:10.1016/j.annepidem.2011.09.001

  1. Wang R, Zhao J, Meitner MJ, Hu Y, Xu X. Characteristics of urban green spaces in relation to aesthetic preference and stress recovery. Urban Forestry & Urban Greening. 2019;41:6-13. doi:10.1016/j.ufug.2019.03.005

  1. Jiang X, Larsen L, Sullivan W. Connections Between Daily Greenness Exposure and Health Outcomes. IJERPH. 2020;17(11):3965. doi:10.3390/ijerph17113965

  1. Astell-Burt T, Feng X. Association of Urban Green Space With Mental Health and General Health Among Adults in Australia. JAMA Netw Open. 2019;2(7):e198209. doi:10.1001/jamanetworkopen.2019.8209

  1. Wang R, Zhao J, Meitner MJ, Hu Y, Xu X. Characteristics of urban green spaces in relation to aesthetic preference and stress recovery. Urban Forestry & Urban Greening. 2019;41:6-13. doi:10.1016/j.ufug.2019.03.005

  1. Giles-Corti B, Broomhall MH, Knuiman M, et al. Increasing walking. American Journal of Preventive Medicine. 2005;28(2):169-176. doi:10.1016/j.amepre.2004.10.018

  1. Schipperijn J, Bentsen P, Troelsen J, Toftager M, Stigsdotter UK. Associations between physical activity and characteristics of urban green space. Urban Forestry & Urban Greening. 2013;12(1):109-116. doi:10.1016/j.ufug.2012.12.002

  1. Urban Green Spaces and Health. World Health Organization; 2016. https://iris.who.int/items/fa2eb466-1013-4773-a999-f029730d7898

  1. Annerstedt Van Den Bosch M, Mudu P, Uscila V, et al. Development of an urban green space indicator and the public health rationale. Scand J Public Health. 2016;44(2):159-167. doi:10.1177/1403494815615444

  1. Kaczynski AT, Besenyi GM, Stanis SAW, et al. Are park proximity and park features related to park use and park-based physical activity among adults? Variations by multiple socio-demographic characteristics. Int J Behav Nutr Phys Act. 2014;11(1):146. doi:10.1186/s12966-014-0146-4

  1. Elmendorf W, Willits F, Sasidharan V. Urban Park And Forest Participation And Landscape Preference: A Review Of The Relevant Literature. AUF. 2005;31(6):311-317. doi:10.48044/jauf.2005.040

  1. RAND Corporation, Derose KP, Han B, Williamson S, Cohen DA. Racial-Ethnic Variation in Park Use and Physical Activity in the City of Los Angeles. J Urban Health. 2015;92(6):1011-1023. doi:10.1007/s11524-015-9994-8

  1. White MP, Elliott LR, Wheeler BW, Fleming LE. Neighbourhood greenspace is related to physical activity in England, but only for dog owners. Landscape and Urban Planning. 2018;174:18-23. doi:10.1016/j.landurbplan.2018.01.004

  1. Astell-Burt T, Navakatikyan MichaelA, Feng X. Why might urban tree canopy reduce dementia risk? A causal mediation analysis of 109,688 adults with 11 years of hospital and mortality records. Health & Place. 2023;82:103028. doi:10.1016/j.healthplace.2023.103028

  1. Lega C, Gidlow C, Jones M, Ellis N, Hurst G. The relationship between surrounding greenness, stress and memory. Urban Forestry & Urban Greening. 2021;59:126974. doi:10.1016/j.ufug.2020.126974

  1. Jimenez MP, Elliott EG, DeVille NV, et al. Residential Green Space and Cognitive Function in a Large Cohort of Middle-Aged Women. JAMA Netw Open. 2022;5(4):e229306. doi:10.1001/jamanetworkopen.2022.9306

  1. Peralta M, Viscioni G, Melo X, et al. Does the Installation or the Improvement of Existing Outdoor Parks Increase Physical Activity Levels? A Systematic Review. Sports. 2023;11(11):221. doi:10.3390/sports11110221

  1. Van Cauwenberg J, Nathan A, Barnett A, Barnett DW, Cerin E, the Council on Environment and Physical Activity (CEPA)-Older Adults Working Group. Relationships Between Neighbourhood Physical Environmental Attributes and Older Adults’ Leisure-Time Physical Activity: A Systematic Review and Meta-Analysis. Sports Med. 2018;48(7):1635-1660. doi:10.1007/s40279-018-0917-1

  1. Finlay J, Esposito M, Li M, et al. Neighborhood active aging infrastructure and cognitive function: A mixed-methods study of older Americans. Preventive Medicine. 2021;150:106669. doi:10.1016/j.ypmed.2021.106669

  1. on behalf of the Council on Environment and Physical Activity (CEPA) – Older Adults working group, Barnett DW, Barnett A, Nathan A, Van Cauwenberg J, Cerin E. Built environmental correlates of older adults’ total physical activity and walking: a systematic review and meta-analysis. Int J Behav Nutr Phys Act. 2017;14(1):103. doi:10.1186/s12966-017-0558-z

  1. Sylvers DL, Hicken M, Esposito M, Manly J, Judd S, Clarke P. Walkable Neighborhoods and Cognition: Implications for the Design of Health Promoting Communities. J Aging Health. 2022;34(6-8):893-904. doi:10.1177/08982643221075509

  1. De Keijzer C, Tonne C, Basagaña X, et al. Residential Surrounding Greenness and Cognitive Decline: A 10-Year Follow-up of the Whitehall II Cohort. Environ Health Perspect. 2018;126(7):077003. doi:10.1289/EHP2875

  1. Wan C, Shen GQ, Choi S. Underlying relationships between public urban green spaces and social cohesion: A systematic literature review. City, Culture and Society. 2021;24:100383. doi:10.1016/j.ccs.2021.100383

  1. Mowen AJ, Rung AL. Park-based social capital: are there variations across visitors with different socio-demographic characteristics and behaviours? Leisure/Loisir. 2016;40(3):297-324. doi:10.1080/14927713.2016.1253178

  1. Jennings V, Rigolon A, Thompson J, Murray A, Henderson A, Gragg RS. The Dynamic Relationship between Social Cohesion and Urban Green Space in Diverse Communities: Opportunities and Challenges to Public Health. IJERPH. 2024;21(6):800. doi:10.3390/ijerph21060800

  1. Finlay J, Franke T, McKay H, Sims-Gould J. Therapeutic landscapes and wellbeing in later life: Impacts of blue and green spaces for older adults. Health & Place. 2015;34:97-106. doi:10.1016/j.healthplace.2015.05.001

  1. Clarke M, Cadaval S, Wallace C, et al. Factors that enhance or hinder social cohesion in urban greenspaces: A literature review. Urban Forestry & Urban Greening. 2023;84:127936. doi:10.1016/j.ufug.2023.127936

  1. Zijlema WL, Triguero-Mas M, Smith G, et al. The relationship between natural outdoor environments and cognitive functioning and its mediators. Environmental Research. 2017;155:268-275. doi:10.1016/j.envres.2017.02.017

  1. Jones A, Ali MU, Mayhew A, et al. Environmental risk factors for all-cause dementia, Alzheimer’s disease dementia, vascular dementia, and mild cognitive impairment: An umbrella review and meta-analysis. Environmental Research. 2025;270:121007. doi:10.1016/j.envres.2025.121007

  1. Hwang SA, Singhvi A, Patil L, et al. A Comprehensive Systematic Review and Meta-Analysis to Unravel the Noise-Dementia Nexus. Public Health Rev. 2025;46:1607355. doi:10.3389/phrs.2025.1607355

  1. Diener A, Mudu P. How can vegetation protect us from air pollution? A critical review on green spaces’ mitigation abilities for air-borne particles from a public health perspective - with implications for urban planning. Science of The Total Environment. 2021;796:148605. doi:10.1016/j.scitotenv.2021.148605

  1. Van Renterghem T, Forssén J, Attenborough K, et al. Using natural means to reduce surface transport noise during propagation outdoors. Applied Acoustics. 2015;92:86-101. doi:10.1016/j.apacoust.2015.01.004

  1. Abhijith KV, Kumar P, Gallagher J, et al. Air pollution abatement performances of green infrastructure in open road and built-up street canyon environments – A review. Atmospheric Environment. 2017;162:71-86. doi:10.1016/j.atmosenv.2017.05.014

  1. Choi EY, Lee H, Chang VW. Cumulative exposure to extreme heat and trajectories of cognitive decline among older adults in the USA. J Epidemiol Community Health. 2023;77(11):728. doi:10.1136/jech-2023-220675

  1. Aram F, Higueras García E, Solgi E, Mansournia S. Urban green space cooling effect in cities. Heliyon. 2019;5(4):e01339. doi:10.1016/j.heliyon.2019.e01339

  1. Bowler DE, Buyung-Ali L, Knight TM, Pullin AS. Urban greening to cool towns and cities: A systematic review of the empirical evidence. Landscape and Urban Planning. 2010;97(3):147-155. doi:10.1016/j.landurbplan.2010.05.006

  1. Chen H, Kwong JC, Copes R, et al. Exposure to ambient air pollution and the incidence of dementia: A population-based cohort study. Environment International. 2017;108:271-277. doi:10.1016/j.envint.2017.08.020

  1. Cariñanos P, Casares-Porcel M. Urban green zones and related pollen allergy: A review. Some guidelines for designing spaces with low allergy impact. Landscape and Urban Planning. 2011;101(3):205-214. doi:10.1016/j.landurbplan.2011.03.006

  1. Abhijith KV, Kumar P, Gallagher J, et al. Air pollution abatement performances of green infrastructure in open road and built-up street canyon environments – A review. Atmospheric Environment. 2017;162:71-86. doi:10.1016/j.atmosenv.2017.05.014

  1. Crous-Bou M, Gascon M, Gispert JD, et al. Impact of urban environmental exposures on cognitive performance and brain structure of healthy individuals at risk for Alzheimer’s dementia. Environment International. 2020;138:105546. doi:10.1016/j.envint.2020.105546

  1. Besser LM, Chang LC, Evenson KR, et al. Associations Between Neighborhood Park Access and Longitudinal Change in Cognition in Older Adults: The Multi-Ethnic Study of Atherosclerosis. JAD. 2021;82(1):221-233. doi:10.3233/JAD-210370

Back to the top or recommendation list

SIGN UP FOR OUR MAILING LIST


Contact us at contact@gatewayexposome.org

The GECC is funded by the National Institute on Aging (NIA) U24AG088894.