Introduction

The twenty-first century is characterized by the ageing of the world’s population. Between 2017 and 2050, the number of adults aged 60 years and over is projected to more than double from 962 million to approximately 2.1 billion [1]. Considering this important demographic shift, factors that support healthy ageing are increasingly important. Healthy ageing is defined by the WHO as “the process of developing and maintaining the functional ability that enables wellbeing in older age” [2] and may be partially determined by environmental factors [3,4,5,6]. In this context, healthy ageing may be supported by exposure to outdoor greenspace.

Recent studies have found that long-term exposure to greenspace was associated with improved health, including better self-perceived general health and mental health, lower risk of type II diabetes, and decreased mortality [7, 8]. The association between greenspace and health may be modified by age as some studies observed that the association was stronger among older adults [9, 10]. However, the evidence for an association between long-term exposure to greenspace and health among the older population has not been systematically reviewed yet. So far, one systematic review summarized the evidence for an association between exposure to nature-based solutions and urbanization-related health risks in the older population, including only a selection of health outcomes [11].

Therefore, the aim of this study was to systematically review the existing evidence on the association between long-term exposure to outdoor greenspace and healthy ageing. We focused solely on outdoor residential greenspace, since the evidence of the association between indoor exposure to nature and health in older adults had been reviewed recently [12].

Methods

Selection Criteria

The selection criteria applied for this review were the following: (a) the article was available in English and concerned an original research article of an observational study. Review articles, experimental studies and qualitative studies were excluded. (b) The association with a health outcome at older age was reported. We included studies on any health outcome, including physical and mental health, wellbeing, overall quality of life and outcomes concerning functioning such as cognitive and physical function. We also considered physical activity as an indicator of physical functioning as this outcome is strongly interrelated with mobility [13, 14]. However, studies on mortality and longevity were excluded, because these outcomes could not provide information on the quality of life at older age (i.e. healthy ageing). Additionally, the evidence for the association with residential greenspace exposure and mortality has been recently systematically reviewed [7]. (c) At least one of the exposures was a quantified measure of long-term greenspace exposure. Studies were excluded if the exposure variable did not assess long-term exposure, was not quantified (for instance, binary variables such as presence or absence of a park without an indication of the objective or perceived distance) or was not assessed as separate predictor of health (e.g. a land use mix, a built environment index or the percentage of green and open areas together). We excluded articles on gardening, since a recent review is available on the evidence of the association between gardening and health [15, 16]. (d) The study population consisted of older adults. While older adults are frequently defined by age 60 and older [1], studies have applied different definitions of older age. In this case, we accepted the definition of older age as given by the study. In addition, as healthy ageing is a process over time and as the age-related decline in body functions may start in middle adulthood, we included studies that include both middle-aged and older adults. Middle adulthood (or midlife) is commonly defined as starting at 40 or 45 years old. Studies including study populations with a minimum age between 35 and 40 years were considered if only a small proportion (< 10%) of the study population was < 40 years old. Last, studies were excluded if the study population only consisted of patients or other non-healthy populations (e.g. populations with dementia at baseline).

Search Strategy

We first searched freely to collect relevant articles to construct an extensive list of keywords that capture articles within the scope of our review (i.e. exposure to greenspace and healthy ageing). We created the search terms based on a combination of greenspace keywords, health keywords and age keywords. The greenspace keywords contained green space(s) or greenspace(s), natural environment(s), outdoor environment(s), natural outdoor environment(s), natural space(s), outdoor environment(s), open space(s), park(s), greenness, green area(s), vegetation, tree cover, VCF, land cover, greenery, garden(s), residential green, nature-based solution(s), nature contact and contact with nature. The health keywords included health, healthy, wellbeing or well-being, quality of life, disease(s), morbidity, life expectancy, longevity, mortality, survival, cognitive decline, physical functioning, dementia, frailty, deterioration, impairment and activity/activities of daily living. The age keywords encompassed older, oldest, oldest-older, middle-aged, mid-aged, mid-older, middle-to-older, elderly, senior, aged, aging, ageing, life course, geriatric, age friendly or age-friendly, retired and retirement. To narrow down the search, we excluded the keywords child, children, adolescent(s), youth and gardening.

We used this list of search terms to extract articles from both PubMed (National Library of Medicine) and Scopus. The search was conducted on March 22, 2019. The retrieved articles were inserted into the online tool Rayyan [17], which facilitates the management of the search results. Articles that clearly did not meet the selection criteria were first excluded based on the screening of the title and the abstract. Afterwards, the full texts of the remaining articles were analysed by two reviewers (CK and MB) independently to decide which articles met the selection criteria and were to be included in the review. The reviewers discussed articles that were subject of disagreement to reach a consensus on the included articles. Last, we scanned through the references of the selected articles to identify relevant additional articles.

Quality Assessment

First, we extracted the following information from the selected studies: study design, study population, sample size, exposure assessment, outcome assessment, main results and additional findings, statistical analysis, covariates and other relevant information (Table 1 and Table S1). Next, we composed 12 quality criteria to score each study (Table S2). These criteria were adapted from previous reviews on health benefits of greenspace exposure [57,58,59]. A score of 0, 1 or 2 points was assigned for each criterion. The two reviewers scored the articles independently and discussed points of disagreement to harmonize the scores. For each study, the points obtained for all criteria were summed up and converted into a percentage of the maximum score. The quality of the study was assessed based on this percentage with ≥ 81% as very good quality, between 61 and 80% as good quality, between 41 and 60% as fair quality, between 21 and 40% as poor quality and ≤ 20% as very poor quality [59, 60].

Table 1 Main characteristics of selected articles on long-term greenspace exposure and mental health, cognitive function and physical capability at older age
Full size table

Evaluation of the Evidence

To evaluate the strength of the overall evidence for the relationship between the exposure and outcome, we classified the evidence per outcome as (a) sufficient, (b) limited, (c) inadequate evidence for an association or (d) evidence for lack of association. The level of evidence was rated based on the guidelines suggested by the International Agency for Research on Cancer (IARC) and adapted by other reviews similar to this review [60, 61]. Sufficient evidence was considered if most of the studies, including good quality studies, observed an association. Limited evidence included several independent good quality studies that reported an association, but the evidence was not yet conclusive. Inadequate evidence was considered if the association was reported by one or more studies, but the studies were of insufficient quality, there were an inadequate number of studies, the findings lacked consistency and/or there was a lack of statistical power. Last, evidence for lack of association included several good quality studies that consistently observed no relationship.

Results

Study Selection

Using our search terms, 2704 unique articles were found by searching PubMed and Scopus (Figure 1). A total of 2489 articles were excluded based on the title and the abstract. Of the remaining 215 full texts, 50 articles were found to meet the selection criteria and were included in the review. Based on the reference lists of these articles, we identified nine additional articles to be included in the review, resulting in a total of 59 articles.

Study Characteristics and Findings

The majority of the identified studies on the association between long-term greenspace exposure and health at older age were of cross-sectional design (N = 44). The remaining studies were of ecological (N = 1) and longitudinal design (N = 14). The studies were conducted in 15 different countries, with 18 studies from the UK and 15 from the USA. Around half of the studies focused solely on the older population (i.e. minimum age was 60 years or older), while the other half also included middle-aged adults. All studies included an objective assessment of long-term residential exposure to greenspace, using spatial data obtained from satellite images, land use or cover maps, administrative data or street view images. Most studies (N = 37) were based on only one indicator of greenspace exposure; only seven studies provided various indicators of exposure to greenspace (e.g. by obtaining different vegetation indices or by using different sources of spatial data) [22, 32, 37, 48, 50, 52, 62], four considered the quality of the greenspace [38, 62,63,64], and three included the use of greenspace in the analyses [26, 65, 66].

The included studies investigated the association between long-term greenspace exposure and various health outcomes. We categorized the studies by health outcome and presented the main characteristics of each study in Tables 1, 2 and Figure 1 (additional information is presented in Table S1). The studies included the following six categories of health: mental health, cognitive function, physical capability, morbidity, cardiometabolic risk factors and perceived wellbeing. The categories were based on the biomarkers of healthy ageing as proposed by an expert panel [82, 83]. If a study reported the association between the exposure to greenspace for more than one health outcome falling into different categories, the study was rated repeatedly for each corresponding category.

Table 2 Main characteristics of selected articles on long-term greenspace exposure and morbidity, cardiometabolic risk factors and perceived wellbeing at older age
Full size table
Fig. 1
figure 1

Selection process of the articles

Full size image

The results of the quality assessment are presented in Table S3. A short description of the characteristics, results of the studies and evaluation of the evidence is given below per outcome category.

Mental Health

We identified 12 studies on the association between long-term greenspace exposure and mental health, including three longitudinal studies [19, 24, 25], eight cross-sectional studies [18, 20, 22, 23, 26,27,28,29] and one ecological study [21]. Most studies focused on depression [21,22,23,24,25, 28, 29], stress [18, 23, 25, 27], and/or anxiety symptoms [24,25,26], mainly assessed with a questionnaire, and in two studies, the outcome was based on the diagnosis of depression (yes/no), obtained from a health administration database [20] or self-reported [19].

Eight out of the 12 studies found that greater long-term exposure to greenspace was associated with a lower risk of stress, depression and anxiety. Furthermore, one study found a non-linear association and three studies did not find any statistically significant association. As the three studies that did not observe any association were considered to be of good quality, we considered the evidence for a beneficial association between greenspace exposure and mental health to be inadequate.

Cognitive Function

There were six articles on cognitive function, including diagnosis with Alzheimer’s disease obtained from a health administration database [20] and assessment of cognitive function by cognitive tests [30,31,32,33,34].

The two longitudinal studies found a beneficial association between greenspace exposure and cognitive decline over the follow-up period [30, 32]. However, the findings of the four cross-sectional studies were mixed; one study found that greater greenspace exposure was associated with lower odds of Alzheimer’s disease [20], but, in contrast, two cross-sectional studies found that higher availability of greenspace was associated with an increased risk of dementia and/or cognitive impairment [33, 34], while another cross-sectional study did not find any association between proximity to park and cognitive function [31]. As the findings were inconsistent and the number of studies low, we considered the evidence for a beneficial association between greenspace exposure and cognitive function at older age to be inadequate.

Physical Capability

The association with physical functioning was assessed in 22 studies, including 17 cross-sectional [35, 38,39,40,41,42,43,44,45,46,47,48, 50,51,52,53,54, 56] and four longitudinal studies [36, 37, 49, 55].

Most studies focused on physical activity (N = 17) assessed by self-reported physical activity levels [35, 36, 39, 40, 42, 44, 46, 48, 50, 52, 53] or objectively measured physical activity using an accelerometer or pedometer [41, 43, 47, 49, 54, 56]. The findings for an association between long-term greenspace exposure and physical activity were mixed; 11 studies observed a beneficial association between greenspace exposure and an outcome of physical activity, but six studies did not. In addition, only four studies were of good quality. Therefore, the evidence for an association between long-term greenspace exposure and physical activity was rated to be inadequate.

The five other studies focused on physical capability assessed by physical tests (e.g. walking speed, grip strength, and timed up and go) [37, 45] or self-reported functional limitations, frailty, or disability [38, 51, 55]. Two studies were longitudinal [37, 55] and three cross-sectional [38, 45, 51]. The two studies of good or very good quality found a beneficial association between long-term greenspace exposure and physical functioning, but the three studies rated of fair quality observed no significant association. Considering these mixed findings and the small number of studies, the evidence for a beneficial association between long-term greenspace and physical capability was considered to be inadequate.

Morbidity

We identified 11 studies on the risk of disease, including eight cross-sectional [42, 63, 67,68,69, 72,73,74] and three longitudinal studies [65, 70, 71]. A wide range of diseases including diabetes [42, 67, 70, 71], skin cancer [68], cardiovascular disease [42, 65, 72,73,74] and cardiocerebral vascular, joint, digestive, endocrine, urological, nervous system and respiratory diseases [63] were assessed. Six studies had objective data on the outcomes [65, 69,70,71,72, 74], while five studies were based solely on self-reported disease [42, 63, 67, 68, 73].

Nine of the 11 studies found a beneficial association between long-term greenspace exposure and the risk of disease among older adults. In addition, one study found a non-linear relationship and one study observed no association at all. Based on the consistent findings of the studies, including six of good or very good quality, we considered there was limited evidence for a beneficial association between long-term greenspace exposure and the risk of disease.

Cardiometabolic Risk Factors

There were nine articles that investigated the association between long-term greenspace exposure and cardiometabolic risk factors, including weight status [42, 50, 54, 75,76,77,78], hypertension [42, 72, 79] and/or cholesterol levels [42]. Only one study had a longitudinal design [77] and the other eight studies used cross-sectional data [42, 50, 54, 72, 75, 76, 78, 79]. Most studies obtained the outcome data from objective measurements [42, 50, 54, 72, 77,78,79], but two studies used self-reported height and weight [75, 76].

The findings were mixed. Among the seven studies on the association between long-term greenspace exposure and weight status, three observed a significant beneficial association, one a non-linear association, and three did not observe a significant association. Considering the association with hypertension, one study observed a beneficial association, one observed a non-linear association, and one did not observe an association with hypertension. Only one study investigated the association with cholesterol levels but did not observe a significant association. Considering the small number of studies and the mixed results, we considered the evidence for an association between greenspace and cardiometabolic risk factors to be inadequate.

Perceived Wellbeing

The association between long-term greenspace exposure and self-perceived wellbeing was assessed in nine cross-sectional studies, including self-rated health status [26, 38, 54, 64, 66, 81], self-reported wellbeing [80] and (health-related) quality of life [51, 62]. Six of the nine studies did not find any significant association, while three found a beneficial association. Only two studies were rated to be of good quality; we therefore considered the evidence for an association between long-term greenspace exposure and perceived wellbeing to be inadequate.

Discussion

Limitations of Available Evidence

Though this review included 59 studies on various health outcomes, we were limited by a small number of studies per health outcome. In addition, the studies were heterogeneous in study design, exposure assessment, statistical methodology and study population samples which complicated comparison and interpretation of the different results. Altogether, conducting a meta-analysis for our reviewed associations was not feasible.

Study Design

Overall, a small part of our reviewed studies had longitudinal design, while majority of the studies were cross-sectional. Reverse causality cannot be ruled out when using cross-sectional data as the outcome may precede the exposure. Furthermore, cross-sectional studies are prone to self-selection bias when, for instance, less healthy adults move to neighbourhoods with more greenspace available. Nevertheless, 16 of the 44 cross-sectional studies took into account the residential history of the participants by selecting only study participants who had not moved recently or by adjusting for length of residency. Longitudinal studies are less prone to reverse causation and self-selection bias and by providing trajectories of health status or incidence of disease over the time, they are more capable of evaluating the effects of greenspace exposure on the ageing process.

Exposure Assessment

All studies included in this review obtained an objective indicator of greenspace exposure, mainly based on satellite-based indices of greenspace or land use or cover maps. In addition, all studies assessed the greenspace exposure at the residential location. However, in several studies, the risk of exposure misclassification could not be ruled out as the residential location was not based on the residential address, but on, for instance, the postcode centroid or administrative area. Furthermore, none of the studies assessed the exposure to greenspace at another location than home, while older adults may also spend a part of their time outside of their direct neighbourhood.

Ideally, to assess the exposure to greenspace, different aspects should be assessed such as the physical and visual access, the actual use and the quality of greenspace. Moreover, the type of vegetation and the richness of biodiversity in greenspaces are potentially relevant. However, among the studies included in this review, most studies only used a single greenspace indicator. Only a few studies included a comparison between various indicators of greenspace exposure or considered the quality or use of a greenspace. Therefore, the type or specific characteristics of greenspace that may be most supportive of healthy ageing are still largely unknown.

Outcome Assessment

The review identified studies on the association of long-term greenspace exposure with a wide range of health outcomes, including outcomes of mental health, cognitive function, physical capability, cardiometabolic risk factors, morbidity and perceived wellbeing. Cognitive function and physical capability are key indicators of healthy ageing [82]. This review identified six studies on cognitive function, of which only two had a longitudinal design. Regarding physical capability, the assessment of locomotor function, strength, balance and dexterity have been proposed to be most indicative of age-related physical capability [82], but this review only identified two studies that measured walking speed and grip strength. The studies on physiological function included in this review focused on cardiometabolic risk factors and the assessed outcomes were weight status, hypertension and cholesterol level. However, several important biomarkers of age-related physiological function such as lung function or glucose homeostasis have not been explored [82]. Additional relevant biomarkers of healthy ageing may be indicators of endocrine function, sensory functions and immune function [82], but we did not identify studies looking at the association of long-term greenspace exposure with these outcomes. Similarly, the potential impact of greenspace exposure on telomere length or other markers of cellular ageing remains as an open question to be evaluated by future studies.

Mechanisms

Long-term exposure to greenspace may be supportive of healthy ageing through various pathways. First, more greenspace in the residential environment could lead to less feelings of loneliness, more social support and improved social cohesion in the neighbourhood [84, 85], which are important contributors to health at older age [3]. Second, greenspace may be a resource for psychological restoration [86]. Exposure to greenspace has been associated with reduced stress [87] and providing the opportunity to restore directed attention [88,89,90], which may benefit cognitive ageing. Third, older adults living in areas with higher access to greenspace have shown higher physical activity levels [39] and a reduced decline in physical activity [36], while physical activity plays a significant role in maintaining functioning and health at older age [91, 92]. Last, increased exposure to greenspace has been associated with lower exposure to environmental stressors such as air pollution, noise and heat [93], which are detrimental to health at older age [70, 94, 95].

Few of the studies in this review conducted formal examination of these potential mechanisms. We were unable to compare these results due to the low number of studies and the heterogeneity in applied statistical methods. Consequently, this review could not provide sufficient information to further understand the pathways (Table S1).

Conclusions

In this review of observational studies on the association between long-term exposure to outdoor greenspace and healthy ageing, we identified 59 studies on outcomes of mental health, cognitive function, physical capability, morbidity, cardiometabolic risk factors and perceived wellbeing at older age. Overall, although the available evidence for a beneficial association between greenspace exposure and the aforementioned outcomes is still limited/inadequate, they are suggestive for the existence of such associations and call for future studies to establish the associations.

Recommendations

The findings of the articles included in this review call for future studies, especially studies that (a) use a longitudinal design that provide insight in the process of ageing; (b) objectively assess healthy ageing by using, for instance, repeated measures of biomarkers of healthy ageing [82]; (c) assess the exposure to greenspace repeatedly over the study period and include various aspects of greenspace exposure; and (d) investigate the underlying pathways for the association between greenspace and health at older age.