Aging is a worldwide phenomenon that impacts both developed and developing countries; China is experiencing rapid and unprecedented aging, with its population aged 60 and over projected to reach 28% by 2040 (World Health Organization [WHO], 2023). Maintaining the health of elderly individuals has become a pressing public health issue, as they face a high burden of non-communicable diseases, disabilities, and inequalities (Liu et al., 2023; Wang and Chen, 2014).

Accumulating findings worldwide demonstrate that human interaction with nature, particularly greenspaces, generates significant physical and mental health benefits (Gascon et al., 2016; Rojas-Rueda et al., 2019; Twohig-Bennett and Jones, 2018). Specifically, exposure to greenspaces is associated with long-term reduction of stress (Thompson Coon et al., 2011), accelerated recovery speed following surgical procedures (Ulrich, 1984), lower risk of type II diabetes (Astell-Burt et al., 2021), cardiovascular disease (Yao et al., 2022), respiratory problems (Orioli et al., 2019), and diminished risk of other chronic diseases (R. Mitchell and Popham, 2008). As individuals age, the incidence of the aforementioned diseases is higher among the elderly population (Ali et al., 2022; De Keijzer et al., 2020). Research has observed that the health benefits arising from exposure to greenspaces are more pronounced in elderly individuals (Astell-Burt et al., 2014; De Vries et al., 2003). Besides, greenspace facilitates interpersonal connections and fosters a feeling of belonging within a community (Ali et al., 2022). This aspect holds immense significance for the well-being of elderly individuals, who face heightened vulnerability to social detachment (Steptoe et al., 2013). The evidence underscores the crucial role of promoting access to greenspaces as a key strategy for improving public health, especially among the aging population.

Furthermore, the potential of greenspaces to mitigate health inequalities, articulated as equigenesis hypothesis, has recently gained research attention (Frumkin et al., 2017; R. J. Mitchell et al., 2015; R. Mitchell and Popham, 2008). The hypothesis postulated that the groups with low socioeconomic status (SES), who typically have less access to alternative resources promoting health, may benefit more from greenspaces, particularly those that are freely accessible (Mitchell and Popham, 2008; Moran et al., 2021).

Due to their straightforward, objective, and comprehensible nature, all-cause mortality has been widely utilized as indicators to assess the overall health and well-being of a population (Cheng et al., 2021; Connolly et al., 2023; Moran et al., 2021). However, existing research has shown inconsistent results in assessing the impact of greenspace. Several researchers suggested an association between higher exposure to greenspaces and decreased all-cause mortality (Markevych et al., 2017; Rojas-Rueda et al., 2019; Twohig-Bennett and Jones, 2018). Nevertheless, a meta-analysis also reported an inconclusive link between greenspace exposure and all-cause mortality (Gascon et al., 2016).

Similarly, the existing evidence for the equigenesis hypothesis is also inconsistent. Some empirical studies support this hypothesis. For example, the effects of greenspace exposure were greater in groups with lower SES on all-cause mortality (R. Mitchell and Popham, 2008), mental health (McEachan et al., 2016; Mitchell et al., 2015; Wang et al., 2022), and violence-related mortality (Moran et al., 2021). However, there is also some evidence that does not support the equigenesis hypothesis, such as the evidence regarding general health (Feng and Astell-Burt, 2017), mental health (Sugiyama et al., 2016), and cardiovascular diseases mortality and life expectancy (Moran et al., 2021).

The variability in the evidence may be due to the variations in the methodologies employed to quantify greenspace across different studies. Although several nationwide research has investigated the health impact of greenspaces (Helbich et al., 2018; Kondo et al., 2020; Schinasi et al., 2019; Villeneuve et al., 2012; Wang and Tassinary, 2019) and the equigenesis hypothesis (Mitchell and Popham, 2008; Moran et al., 2021), most of them were conducted at the aggregate level (e.g., fishnet unit, zip code, census tract) rather than individual level. These aggregate-level studies are prone to ecological fallacy, the bias linking to the assumption that individuals share the same characteristics as their aggregate group, neglecting within-group and individual variations. Thus, such bias may result in a failure of hypotheses or conflicting results (Freedman, 1999; Piantadosi et al., 1988). In most circumstances, the frequency of ecological fallacy can be attributed to the absence of detailed individual-level data at a fine-scale level.

To avoid ecological fallacy, individual-level studies have been recently employed in health-greenspace studies. These studies typically used the straight-line buffer to assess the neighborhood greenspace exposure (Astell-Burt et al., 2021; Bauwelinck et al., 2021; Orioli et al., 2019); the straight-line buffer defines a circular area, which assumes that there are no obstacles and people can walk anywhere. However, in an urban environment, the activity space of an individual is limited by physical boundaries, such as fences, buildings, and rivers. Applying the street-network buffer, which assumes people can only access the area reachable within a street network, can result in more nuanced representations of an individual's surrounding environment (Droin et al., 2023; Ho et al., 2022, 2023). Although street-network buffers can provide a more accurate greenspace exposure, few studies have employed it to investigate the impact of exposure to greenspaces on health outcomes (e.g., mortality, life expectancy). In addition, although some scholars have pinpointed that different measurements of greenspace may be a primary factor in the inconsistency of the evidence (Feng and Astell-Burt, 2017; Wang et al., 2022), no empirical research has compared health outcomes with different measures of greenspace exposure (i.e., aggregate level vs. individual level and straight-line buffer and street-network buffer) using the same dataset at the citywide level.

In recent years, increasing health-related studies have focused on life expectancy loss, which is typically evaluated by years of life lost (YLL) (Cheng et al., 2021; Guo et al., 2013; Moran et al., 2021; Qi et al., 2020). Compared with mortality rates, YLL provides a more comprehensive and informative measure of the health impact, as it accounts for life expectancy loss at the time of death, adjusted for age and sex demographics (Cheng et al., 2021). Thus, YLL is not vulnerable to biases, places more weight on young-age deaths, and has been recognized as a more advanced measure for evaluating the health effects of exposure or intervention (Cheng et al., 2021; Connolly et al., 2023; Lai and Hardy, 1999). Recently, an increasing number of studies used YLL to assess the life expectancy of exposure to urban environment (e.g., air pollution, heat waves, and cold temperature) (Cheng et al., 2021; Moran et al., 2021; Qi et al., 2020). However, no research has examined the effects of greenspace exposure on the life expectancy of elderly population at the individual level in China.

To address the above research gaps, this study employed the individual-level mortality dataset of elderly population in Guangzhou, China, in 2010 and comprehensively explored the relationship between greenspace exposure and life expectancy. The four objectives were as follows: 1) to examine the relationship between neighborhood greenspace exposure and life expectancy at the individual level; 2) to explore whether neighborhood greenspace exposure has higher effects on life expectancy for socioeconomically disadvantaged individuals (equigenesis hypothesis); 3) to investigate the optimal buffer distances of neighborhood greenspace exposure with life expectancy for elderly population; 4) to compare the results from different measures of greenspace exposure (Fig. 1).

Our research extends existing studies in several aspects. First, this is one of the first research to uncover the impact of greenspace exposure on life expectancy at the individual level in a densely populated Chinese context. This study can avoid ecological fallacy and offer more rigorous evidence compared with the studies using aggregate-level data (Hong et al., 2021; R. Mitchell and Popham, 2008; Moran et al., 2021), examining the effect of greenness and equigenesis hypothesis on life expectancy. Second, by investigating the optimal buffer distances, we can explore the optimal distances of greenspace exposure between neighborhood greenspace exposure and elderly life expectancy, thereby providing support and design strategies for age-friendly urban design and management. Third, by comparing different greenspace measurements, we can ascertain whether greenspace measurements are a potential contributor to the inconsistent results and offer suggestions for future greenspace-health research that involves measuring greenspaces.