Neighborhood greenspaces including parks, gardens, and other areas with natural vegetation such as tree-lined streets are integral elements of the socio-exposome, which is the accumulation of environmental exposures and their social determinants from the peri-conceptual period and throughout an individual's life course (Senier et al., 2017). Greenspaces have been associated with multiple health outcomes among children through older adults (Jimenez et al., 2021; Tzoulas et al., 2007; Tzoulas, 2010; Dadvand, 2021), including lower odds of low birth weight and small for gestational age (Zhan et al., 2020), better mental health (Astell-Burt et al., 2014), longer and higher quality sleep (Shin et al., 2020), lower cardiovascular and cerebrovascular disease risk, greater physical activity (Carpenter, 2013; Xiao et al., 2022), higher cognitive functioning (Besser, 2021a, 2021b), and lower risk for Alzheimer's disease and related dementias (Besser, 2021a). They provide opportunities for physical activity, social engagement, stress relief and relaxation, exposure to natural environments with less air pollutants, and exposure to natural chemicals from trees that are postulated to benefit the immune system (e.g., phytoncides). It is through these mechanisms that greenspaces are hypothesized to relate to human health (Besser, 2021a).

Researchers from health, social science, and environmental disciplines typically use objective measures of neighborhood greenspaces derived from satellite-based imagery or maps/street views. For instance, in a review of the literature on greenspaces and Alzheimer's disease-related outcomes, 21 of the 22 studies used objective (versus perceived) greenspace measures such as the validated normalized difference vegetation index (NDVI) (Besser, 2021a; Rhew et al., 2011). The NDVI, which is calculated from satellite imagery, describes the greenness or amount of healthy vegetation in an area based on light reflectance from plants, with more positive, higher scores indicating greener environments (range: −1 to +1) (NASA Earth Observatory, 2018). Tree canopy cover measures (i.e., amount of ground coverage from leaves, stems, needles, and branches of trees) are similarly derived from satellite imagery (Coulston et al., 2012). Public park space data ascertained from maps or public/licensed data sources can be used to calculate measures such as the number of neighborhood parks or the percentage of the neighborhood composed of park space (Besser et al., 2020, 2021b). Walking audit tools such as the Senior Walking Environment Audit Tool or Wisconsin Assessment of the Social and Built Environment have also been used to capture neighborhood environmental features including presence of parks/playgrounds and shade trees, and often employ two raters to increase reliability of the measures (Malecki et al., 2014; Herbolsheimer et al., 2020; Poortinga et al., 2017). Increasingly, studies are developing detailed measures of greenspace from virtual street views (e.g., Google), such as amounts of grass, trees, and low-lying vegetation at the perspective of people on the ground. Virtual street view techniques have been found to have low to high validity depending on the study (compared to manual review) (Lu, 2019; Vanwolleghem et al., 2014). Other studies have derived various greenspace indices based on more than one factor (e.g., single measure combining data on Normalized Difference Vegetation Index, percent park space, and distance to nearest park) (Slawsky et al., 2022).

While objective measures may provide a more accurate quantitative representation of greenspace compared to self-reported measures, they have some limitations. NDVI and tree canopy cover measures do not capture types of greenspaces (e.g., tree lined streets with sidewalks versus parks) that may be more likely to influence health, and values for these measures will vary over the year and differ across regions, which is often not accounted for in analyses. Geospatial/map data on parks are often not readily available longitudinally, which limits research particularly for regions that experience frequent development resulting in new parks or removal of greenspace. Walking audits are time and resource intensive, are not typically conducted longitudinally, and while they are based on visual assessment are still subject to observer bias. Virtual street view greenspace measures have the benefit of providing rich data on types of greenspaces at the individual's perspective, but like other measures have the limitations of not being systematically assessed longitudinally or by season due to lack of data. Lastly, while composite greenspace indices provide a single measure to describe neighborhoods, which can simplify analyses and help account for high correlation among multiple greenspace measures, they reduce the specificity and translatability of findings to future interventions. In addition to the limitations noted above, objective data on greenspace such as satellite imagery are not readily available prior to the 1980s, and other greenspace data on parks and from street views are available for even fewer years. Thus, while tracing residential history can assist in developing longitudinal greenspace measures for decades prior, it currently cannot be used to assess greenspace exposure in childhood and early adulthood for existing cohorts of older adults.

Perceived measures of greenspace availability, quality, and use provide new perspectives compared to objective measures and also have been associated with health outcomes including but not limited to physical activity, hypertension, and diabetes (Liu et al., 2023; Tabatabaie et al., 2019; Orstad et al., 2017). Unlike objective measures, self-reported measures can detail the varied uses of neighborhood greenspaces (e.g., time of day used, time spent in, greenspace types used, activities in greenspaces), quality and preferences for those spaces (e.g., preferred greenspace types and activities/programming, safety and aesthetics), and perceived access to nearby greenspaces. In some studies, perceived but not objective measures of greenspace have been associated with physical activity, suggesting that greenspace perceptions are just as important to positive health behaviors and outcomes (Tabatabaie et al., 2019; Orstad et al., 2017). Altogether, while perceptions of greenspace may vary between individuals depending on factors such as culture, preference, and neighborhood context, it has been suggested that they may ultimately capture greenspace in ways more closely aligned with greenspace quality than objective measures (Liu et al., 2023; Jones et al., 2009; Leslie et al., 2010).

The type and number of perceived greenspace items measured have varied depending on the research study, and no standardized measures or questionnaires are broadly in use (Loder et al., 2020). While many studies are based on a few select questions such as the perceived presence of street trees or the amount of greenspace in the neighborhood (Zijlema et al., 2017; Saelens et al., 2003; Knobel et al., 2021), more extensive surveys have been conducted. For instance, the 2017 Greenspace Use and Attitudes Survey asked 13 questions including distance to the nearest greenspace, frequency of visiting greenspaces, and expectations and perceptions of the greenspaces for residents in Scotland (Greenspace Scotland, 2017). Four greenspace questions from the widely disseminated Neighborhood Environment Walkability Scale (NEWS) have been shown to have moderate to strong test-retest reliability (correlation coefficients: walking proximity to park = 0.67; trees along the streets = 0.63; tree cover or canopy along sidewalks = 0.52; and attractive natural sights in neighborhood = 0.59 (Sallis, 2022a)). However, comparisons between perceptions of the presence of neighborhood parks and tree lined streets measured in NEWS and objective measures calculated using GIS have suggested weak correlations (r = −0.23 and r = 0.06, respectively) (Adams et al., 2009). Overall, the evidence for the reliability and validity of self-reported greenspace measures and instruments remains limited and there is a lack of assessments of perceived greenspace beyond the present neighborhood environment.

Research has primarily assessed greenspaces for an individual's current residential neighborhood. Given the paucity of standardized greenspace questionnaires more generally (Loder et al., 2020), it is unsurprising that studies have yet to focus on perceived greenspace measures from a life course perspective (Halfon et al., 2018). A systematic review of 59 studies that examined longitudinal exposure to greenspace and healthy aging outcomes found that all long-term exposure measures were objective and not self-reported/perceived (de Keijzer et al., 2020). As noted above, objective measures are limited in that historic maps and satellite imagery have limited availability (e.g., USGS satellite imagery was captured annually starting in the 1980s) and thus cannot be linked to the early to midlife residential addresses for many older adults. The exposure to greenspaces over the life course including those earlier periods may have a cumulative and more potent association with health outcomes compared to a single point in time or in later life. In addition, greenspace exposures during particular life stages may be more strongly associated with human health. Thus, experts on the research intersections between health and place have called for a life course approach to studying greenspace and health associations (Pearce et al., 2016). To date, there are no known standalone instruments designed for older adults to self-report perceptions on their neighborhood greenspaces from various time points extending back to childhood.

This study fills this gap by providing such a questionnaire that can be used and adapted to survey older adults regarding their perceptions of greenspaces and other key individual and neighborhood-level characteristics across the life course. The Life Course Sociodemographics and Neighborhood Questionnaire (LSNEQ) was designed as part of a National Institute of Aging K01-funded study (AG063895) to allow examination of associations between perceived neighborhood greenness and park access in early-, mid-, and late-life and brain health measures (i.e., cognition and magnetic resonance imaging). Lack of objective data with which to assess neighborhood greenspaces during earlier periods when studying older adults provided the impetus to design the LSNEQ. In addition, the questionnaire provides a resource for researchers that do not possess time/personnel and expertise in mapping software/geographic information systems (GIS) with which to derive objective measures of the neighborhood environment including greenspace. In this paper, we aimed to: 1) describe the LSNEQ development, items, and index measures, 2) assess the internal consistency and test-retest reliability of the LSNEQ indices, 3) examine differences in LSNEQ index scores depending on participant characteristics (i.e., age, sex, racialized group, education, and location), and 4) investigate whether LSNEQ indices are associated with self-reported neighborhood walking among older adults.