A growing body of evidence shows a relationship between levels of green space in the local neighbourhood and people’s health and wellbeing, especially for low-income and deprived urban or suburban populations [1
]. Studies in Japan have shown that access to forest environments can promote lower concentrations of cortisol, lower pulse rate and blood pressure, greater parasympathetic nerve activity and lower sympathetic nerve activity compared to city environments [4
]. These studies suggest that green space may offer opportunities to buffer or mitigate health outcomes for urban populations. Recent research in the United Kingdom (UK), using objective, GIS-based measures of the amount of green space in the residential environment and biomarkers of stress combined with self-report measures, has shown a significant association between the amount of green space in the environment and stress levels for a deprived urban population not in work [7
Such findings offer evidence that underlines the potential importance of access to green space when developing future strategies for healthier urban environments such as the Scottish Government’s Place Standard [9
]. Studies have attempted to assess how much green space is needed for healthy functioning of a city’s ecological systems [10
] and to consider the role of ecosystems services for human health [11
]. However, questions remain regarding the quantity and types of green space within a community which are associated with significant differences in stress levels and with human health more generally [12
In a study of green space quality and quantity in relation to self-reported general health and certain mental health outcomes, in four large Dutch cities, Van Dillen et al.
] showed that quality measures may add predictive value to quantity measures. Nonetheless, quantity and quality were related to a substantial degree. However, the study also identified the importance of street trees and their contribution to views of greenery, found to be associated with health outcomes. This suggested that measures of green space relevant to health in urban areas should include not just dedicated recreational space such as public parks and playing fields but also other types of green space, such as road verges. A study of allotment gardening by van den Berg and Custers [14
], using salivary cortisol as a measure of stress, found that a period of gardening led to greater levels of stress reduction than a restful indoor task (e.g., reading), adding support to evidence that access to a garden or allotment offers health benefits in general, and benefits for buffering stress [15
] in particular (An allotment is a plot of land rented by an individual for growing vegetables or flowers. Provision of groups of allotments, often owned by urban authorities, is commonplace in many parts of Europe and can be an important green space and gardening resource for urban dwellers.).
A further consideration is the area around a person’s home in which the contribution of green space may be significant with respect to their health. Maas et al.
] used a radius of 1 km and 3 km, respectively, around people’s homes to study the effect of the quantity of green space within the living environment on a range of physical and mental health measures. They found that green space close to home is more important, although a similar study on green space as a buffer between stressful life events and health only found an effect within the wider area of 3 km [16
]. However, more local green space measures have also been shown to be relevant. The study by van Dillen et al.
] measured the quantity of green space within an area of radius 500 m around participants’ homes and identified links between green space quantity and self-reported health. The European Common Indicator of availability of local public open areas does not set a target and is not specifically focused on green space, but is based on the percentage of citizens living within 300 m from a public open area of minimum size 0.5 hectares [17
]. This is considered the approximate distance within a 5-min walk from home, taking into account varying walking speeds of different age groups and barriers such as main roads or indirect routes that may add time to any distance walked.
While useful as a basis for research, such measures of proximity of green space do not necessarily equate with visibility, perceptions or use of that space for local residents. Research in Scotland has shown that, for many people, regular green space use is not limited to locations within 300 m from home [18
] and perceptions of green space proximity and accessibility may relate to what is perceived to be part of the neighbourhood or available for community use, rather than objectively measured distance [19
]. On the issue of green space visibility, the scientific literature on aspects of stress relief and attention restoration [20
] offered by green space suggests that simply viewing green space may offer a range of health benefits [21
]. It is important therefore to consider whether merely viewing green space from one’s home or street is sufficient to be associated with lower stress levels, or whether visiting green space is necessary to gain significant benefit. Equally, it is important to understand whether access to a private garden or allotment offers sufficient green space experience to be associated with lower perceived stress levels, as van den Berg and Custers [14
] suggest, compared with access to a park or other publicly accessible open space.
A relevant consideration in studying links between green space and health is the former’s potential role as an environment conducive to physical activity or social wellbeing. The importance of physical activity for health, and the role of physical environment in supporting activity, has been understood for some time and underlined in recent research reviews [23
]. Recognition of the importance of social capital for health, both physical and mental, is a more recent phenomenon [24
], with social isolation a particular health risk for older people [25
]. Heinrichs et al.
] have shown how social contact can have positive effects on mood and stress level. Choenarom et al.
] have demonstrated the role of a sense of belonging and social support for perceived stress and its links to depression and [30
] demonstrated an inverse relationship between a sense of belonging and perceived stress.
A considerable body of research has reported on environmental support for physical activity, and walking in particular [31
]. However, evidence shows that socio-economic deprivation also plays an important role in such relationships between green space and health. Lachowycz and Jones’s study of a large UK sample [32
] confirmed an association between green space access and reduced cardiovascular mortality found previously [1
] but only amongst the most socio-economically deprived groups. Although a relationship between green space access and walking was found for all areas, the authors found no evidence that recreational walking explained the relationship between green space access and mortality. The authors suggest that the relationship may be explained by mediators other than walking, such as psychosocial factors. Maas et al.
] studied the relationship between lower morbidity and a green living environment in The Netherlands and found the relationship was strongest for anxiety disorder and depression and stronger for children and people of lower socio-economic status. Another UK study [34
] identified significant associations between reported access to, and better quality of, green space and reduced psychological distress in a deprived urban population. Such evidence is supported by a European epidemiological study [3
] of associations between mental wellbeing and people’s level of financial strain; it found that socioeconomic inequality in mental wellbeing was 40% narrower among respondents reporting good access to green space, compared with those with poorer access.
One way in which good access to green space may contribute to reduced health inequalities in income-deprived communities is through frequency of and/or time spent in outdoor activities, which may in turn support social engagement and thereby link to wellbeing [18
]. In a Dutch study, de Vries et al.
] found an association between streetscape greenery and perceived social cohesion at the neighborhood scale. Maas et al.
] considered loneliness, social support, and contact with neighbours and friends in the neighbourhood as potential mediators of links between green space quantity and self-reported health outcomes. They found, overall, that people with more green space in their living environment felt less lonely and less lacking in social support, but they did not report having more contact with neighbours or friends in the neighbourhood or receiving more social support in practice. The authors suggest that the relationship between green space and social contacts has more to do with the fact that green spaces can strengthen a sense of community via residents developing their attachment to place and place identity, than with any greater contact with neighbours.
In their review of green space links with health, Hartig et al.
] emphasise that relationships between social wellbeing and greenspace are complex and not easily explored via experimental research. People’s social wellbeing may be detrimentally affected by green and open space in which they feel unsafe or where other people engage in unwelcome or anti-social behaviour. This may be particularly true for areas of socio-economic deprivation. There is also considerable evidence that women experience fears for personal safety differently from men and that this influences how green space is perceived and used [38
], which may in turn reflect differences by sex in the associations between green space and health [41
]. Jiang et al.
] have shown differences between men and women in relationships between urban tree cover and stress levels and Roe et al.
] found differences by sex in patterns of stress associated with local green space in deprived urban areas in Scotland.
Drawing on the above research, our study attempts to investigate what level or type of availability of green space is associated most strongly with any difference in perceived stress levels, and in general health. It also explores potential pathways for explaining links between green space and stress levels. As the association between green space and stress or mental wellbeing has been shown to be strongest in low socio-economic groups [2
] or populations suffering financial strain [3
], as described above, our study has focused on economically deprived urban populations.
The aims of the study are to investigate:
What measures of quantity or use of local green space are associated with differences in stress levels in deprived urban communities?
Are physical activity or social wellbeing variables also associated with differences in stress levels in deprived urban communities, and how do these relate to green space variables?
Are any relationships between quantity or use of local green space and stress also found for general health as an outcome?
Do physical activity or social wellbeing variables play a role in predicting general health?
Based on the literature suggesting links between access to local green space and health, particularly levels of stress, in economically deprived communities, we explored potential relationships in our study communities’ data. The sample came from four different urban areas in two of Scotland’s central belt cities−a study across and within cities, which is innovative in design. Although all were in areas of high deprivation, based on Carstairs Index for UK census wards, each site had varying levels of green space and access to private gardens or allotments. We found that there were significant differences by community in levels of self-reported stress and general health, and that measures of green space access were significantly associated with these outcome measures in the majority of cases. Although perceived stress and general health are correlated (p < 0.01) in our sample, the predictors for each outcome, including green space variables, were by and large different in each case.
We consider each of the research questions in turn and then relate our findings to the broader hypotheses on which they were based, as well as the limitations of our study.
4.1. Characteristics of Access to Green Space Associated with Differences in Perceived Stress Levels
We tested whether characteristics of access to and quantity of local green space are associated with differences in levels of perceived stress in deprived urban communities. We found that having a private garden or allotment and the total green area of space in the neighbourhood (including private gardens) are both predictors of variation in stress levels across all four communities. For Community 1, the community with the highest stress levels (significantly so, with a mean over 5 points higher on the 40-point PSS scale than the other three communities sampled), access to a garden or allotment was the single best predictor of stress levels. For the other three communities, the percentage of neighbourhood (datazone) green space was a significant predictor of stress: an increase in green space was associated with a decrease in stress levels. However, it was only at levels of 60% green space or above that the majority of participants were found to have stress levels below the median.
For most of our participants, having high levels of green space in the neighbourhood, including private gardens as well as public parks, roadside trees and grass or other vegetated areas, appears to be a part of the living environment associated with stress mitigation. The community (Community 1) with the highest levels of self-reported stress did not have the greatest levels of area-level socio-economic deprivation according to the Carstairs Index, but had the lowest levels of educational achievement and highest levels of people not in full-time employment. This is reflected in its 2009 SIMD score [47
]. SIMD ranks the 6505 datazones in Scotland, from the most deprived (1) to least deprived (6505), with the most deprived 15% ranked between 1 and 976 [62
]. Community 1 had a median of 482, compared to scores of 577, 990 and 543 for Communities 2, 3 and 4, respectively. SIMD was not included in the analysis as it combines measures of health, environment, social and economic context which were being analysed separately. However, the findings from our study support the particularly deprived SIMD measure for Community 1. The significance for stress levels of having access to a private garden or allotment, rather than overall green space, suggests that a green space that offers the potential for personal or private use is important in such a community.
4.2. Physical Activity or Social Wellbeing Variables Associated with Differences in Perceived Stress Levels
We tested whether physical activity or social wellbeing variables were associated with differences in stress levels and whether this association was stronger than for green space variables. Self-reported physical activity levels did not appear in any models predicting stress. However, the social wellbeing variables of social isolation and place belonging were strong predictors of stress in three out of four deprived communities sampled, and it was only after removing these variables that a significant relationship between percentage green space and stress was found. For the third social wellbeing variable, neighbourhood trust, mean scores before analysis were similar across communities and no significant relationship between neighbourhood trust and stress was found.
Based on these findings and the associated mediation tests, it appears that green space does not contribute to stress mitigation via enhanced levels of physical activity. Results suggest that its contribution to stress mitigation is partly mediated by its contribution to enhancing place belonging
, in the case of green space area (i.e.
, quantity), and to reducing social isolation
, in the case of access to a garden or allotment. This is supported by ethnographic studies in two of the communities studied [18
], showing that green space can offer good opportunities for social contact and for engaging with the local environment, cultural and natural, which enhance a sense of belonging and reduce social isolation.
In Community 1, the only community where social wellbeing variables were not the strongest predictor of stress, the higher level of multiple deprivation indicated by its SIMD score may reflect poorer social capital in the community as a whole, beyond any socio-economic deprivation indicated by the Carstairs Index. However, given the mediation role identified, it may be that social contact associated with allotment use is an underlying contributor to stress mitigation, as Wood et al.
] suggest in their study demonstrating the mental wellbeing and general health benefits of allotment gardening.
4.3. Relationships between Access to Local Green Space and General Health
We tested whether relationships between access to local green space and levels of perceived stress were also found for general health as an outcome. While green space variables appeared as significant in relation to self-reported general health in two of the four communities studied, these were different measures of green space access than those significant for stress.
For one community (Community 3), frequency of green space visits in the winter months of September to March was the best and single predictor of general health. Those who visited at least once a month in winter reported significantly better health than those who did not. Although this community was similar to the others sampled in terms of full-time employment levels, it had the lowest Carstairs index of deprivation and the highest levels of education. We suggest that winter levels of access to green space reflect behaviour by those who access green space all year round, as opposed to those who only visit green space, if at all, in summer months. This association may therefore reflect regular, and comparatively frequent, activity in green space which is beneficial for general health. It may also reflect evidence that exposure to daylight in winter months in northern high latitudes is comparatively more important for general health than in summer months [64
Green space measures were less important for the other three communities in predicting general health. For two Communities (2 and 4), self-reported physical activity levels were the most important predictor of health, accompanied in the model by age and relationship status. The importance of winter visits, probably reflecting regular year-round visits, to green space, as predicting health in Community 3, may be reflected in Communities 2 and 4 by the variable of physical activity levels. It is possible that physical activity is associated with regular visits to green space in all three communities but predicted less strongly by green space visits in Communities 2 and 4 than in Community 3.
For the community with the worst self-reported health (Community 1), a view of green space or hills from the home was a significant variable in a model predominated by physical activity, as well as measures of social isolation and place belonging. This was also the community with highest perceived stress levels, and both our data and SIMD levels suggest this community may be unusual compared to the others under study. Nonetheless, general health is influenced by psychological as well as physiological factors and our findings would support previous findings that a natural view or distant prospect offers psychological relief that may contribute to health [65
], or mitigate ill-health, in deprived communities.
4.4. Different Sub-Groups in the Sample
We tested whether there were sub-groups which differ in their pattern of perceived stress and general health in our sample. Our CHAID analysis showed that the primary sub-groups in our data were clusters based on community level differences.
Community 1 stood out as having both worse stress levels and worse general health, despite comparatively high green space area (mean 61.02%), the second highest level for the communities under study. As indicated in Section 2.2
, and in data such as employment and educations levels in Table 1
, this community has remained an area of very high deprivation despite regeneration efforts over several decades and it may be that the effect of this has led to a lasting stigma of the kind Pearce has identified [66
], which continues to affect both health and social outcomes. For this community, access to green space was important both in predicting lower stress (via access to gardens or allotments) and general health (via views to green space or hills from the home). The explanation may lie in the spatial distribution of green space within the community, with most lying in a large unit to the south of the community. Thus, the adjacency of green space to the homes of participants, compared to the quantity present, may be a significant factor in relation to stress and general health outcomes.
For the other communities in our sample (Communities 2, 3 and 4), the level of physical activity is more important for general health in Communities 2 and 4 than any relationship between green space visits and health, unlike in Community 3.
Despite evidence in the literature and previous studies in similar communities that health outcomes, including perceived stress, vary by sex [8
], this variable was only found in one of our models, contributing at a low level to predictions of perceived stress in Communities 2, 3 and 4. Given that women in our sample were more likely to report higher stress, further research might usefully explore whether there are sub-groups within such populations for whom green space area and/or social wellbeing variables are differentially associated with stress levels.
4.5. Links to Underlying Mechanisms or Pathways
Authors such as Jennings et al.
] have called for more research to investigate the health impacts of the natural environment and, specifically, the cultural role of ecosystem services. Our findings add to understandings of the potential mechanisms or pathways that link quantity and types of access to green space and perceived stress or health levels in deprived urban populations. They suggest the following.
The restorative qualities offered by views of green and natural places may contribute to general health (as exemplified in Community 1 by the role of views to green space and hills) via physiological responses that have been demonstrated in a number of studies [4
] The quantity of green space in the neighbourhood, perhaps including views of green space, appears to contribute to place belonging (as exemplified in Communities 2, 3 and 4) and may thereby also contribute to general health, although this link is more tenuous. The fact that percentage of green space was more strongly associated with perceived stress levels than the frequency of visits to green space suggests that it may be green space experienced in moving about the residential neighbourhood while focused on goals other than visiting green space, that is the important factor here. Such findings support earlier studies showing links between chronic stress and percentage green space or natural environment in the residential area [7
The support that parks, open space and allotments offer for neighbourhood contact and maintenance of community connections, which reduce social isolation and enhance place belonging, may explain the association between the percentage of green space in the neighbourhood and lower perceived stress levels. This confirms earlier findings on links between social cohesion or belonging and residential green space [35
] and supports evidence from neuroscience that sense of place and place identity, in which the social and natural environment have particular roles, are important dimensions for human health [68
]. Future research could usefully distinguish between the value of private gardens versus
shared gardens and allotments, where greater social interaction with neighbours is likely. The mediating role of social wellbeing variables may explain why, in our study, views of green space or hills from the home were not associated with stress, while percentage of residential green space and gardens or allotments were.
The support that parks and open space offer for year-round physical activity and outdoor recreational access, especially in winter months, may explain the relationship between access to green space and general health (as exemplified in Communities 2, 3 and 4). Several studies have shown a positive association between frequency of visits to, and/or time spent in, nearby green space and levels of physical activity [69
], which in turn predict health [23
], although the international evidence for associations between green space and levels of physical activity, remains equivocal, suggesting that the relationship may vary considerably between countries and population sub-groups [37
Of the individual characteristics, the variables associated both with perceived stress levels (in Communities 2, 3 and 4) and general health (in Community 1) were: employment, age, and whether or not there are children under 16 in the household. Unsurprisingly, being in full-time employment was associated with lower perceived stress and better health. However, older age was associated with lower stress but poorer general health. Other studies have shown that older age is associated with an increase in poor health and multi-morbidity, especially among deprived populations [71
] but also that the relationship between the amount of green space and health is stronger among older people [72
], with a significant association demonstrated between a green environment and levels of physical activity for those aged over 60 years [73
]. Future research might usefully explore such interactions further in deprived urban communities. In our study, having children in the household was associated with higher stress but better general health; this latter finding may reflect participants’ younger age where there are children in the household. While caring for children may add to stress in adults, the importance for both parents and children of green space near the home is highlighted in recent research which identifies links between positive birth outcomes and access to green space [74
], as well as between green space and children’s wellbeing [70
]. Again, future research might usefully explore these interactions in deprived communities such as those of our study.
There are a number of limitations to our study. Most importantly, a cross-sectional study is only able to show associations and the direction of causality cannot be determined.
Similar to other research [78
], our analyses indicate that the data were strongly clustered around community location although they were intended to be comparable across different communities. This observation implies that community level characteristics are important in understanding the link between green space and our outcome measures. As our focus was on green space, we did not include measures of housing type, age or density in our analysis. These might have associations with health outcomes and suggest potential future avenues that might also be explored. While our data suggest some findings likely to be generalizable to other deprived urban communities in Scotland and possibly other locations in the UK and Europe, they also clearly indicate that local socio-economic and cultural characteristics have a strong influence on perceived stress and general health in the community.
Our measures of green space quantity were based on the most recent and detailed, systematic measures of various types of green and open space area available at the time of study. These data were derived from existing, street level national mapping (Ordnance Survey MasterMap) and detailed mapping of green spaces as part of an audit of the City of Edinburgh. They are now augmented by a national green space map for Scotland [52
]. Future such mapping is proposed for England and Wales, but is not readily available for other parts of Europe. Nonetheless, the measures of green space used and what is included or excluded may influence the associations with stress or health in ways that have not yet been examined. We know that ward level census area data on percentage of green space cover yields different figures from those considered at geographic units of datazones [79
], and derived units (e.g., 300 m buffers around homes). This reflects the different sources of data (e.g., satellite classification, or detailed scale mapping), the inclusion or exclusion of private gardens, and the different geographies used to summarise spatial data, referred to as the Modifiable Unit Area Problem [80
]. Our analysis used what were deemed the most appropriate and accurate measures for our study at the time. However, it may be that further refinement of what is included or excluded in the green space measures, such as seasonally relevant data, yields further understanding.
Measures of the quality of green space were not included in the analysis but might have added predictive power to any associations found. Similarly, analysis of associations between total green space area excluding gardens and allotments was not undertaken but, along with questions on gardening habits of participants, might have yielded more insight into the role of gardening, in particular, in relation to stress. Mapping of the locations of local green space, including gardens and allotments, reported to be used by respondents would have added further understanding of the relevant “neighbourhood” within which such green space is found and the distances people were prepared to travel to access it. This would also indicate to what extent such locations fall within residential datazones. These are fruitful avenues to explore in future research.
Our study elucidates the relationship between access to green space, perceived stress and general health in economically deprived urban populations. We found that the quantity and nature of access to green space, including specifically access to gardens and allotments and the percentage of all types of green space in the neighbourhood, were significant predictors of stress. For most of our communities sampled, as green space in the neighbourhood increased, levels of perceived stress decreased. The frequency of visits to green space, particularly in winter months, and views of green space from the home were significant predictors of general health, although physical activity was found to be the strongest predictor of general health in half of our sampled communities.
The findings underline the importance of social wellbeing as predictors of stress and, to a lesser extent, general health, offering support for findings by Maas et al.
] and further insight into those of de Vries et al.
]. The variables of social isolation and place belonging were strong predictors of perceived stress in three out of four deprived communities sampled, and of poor general health in the fourth (least healthy) community. There are indications of mediation by social wellbeing in the relationship between local green space and stress levels.
The findings suggest that the contribution of green space to health and wellbeing in deprived communities lies partly in an enhanced sense of place belonging and a reduced sense of social isolation, as well as in offering opportunities to manage or mitigate stress and maintain year-round healthy activity, as indicated by winter visits to green space. In such deprived communities, having access to green space in the neighbourhood may buffer some of the effects of stressors such as unemployment. Factors such as natural space, identity and belonging, and social interaction are all now being taken into account in recently developed public policy towards built and natural environments, such as the Scottish Government Place Standard [9
]. The study results have implications for green space standards in urban planning and housing design. For example, in our least healthy community, experiencing greater levels of stress, having access to a garden or allotment was a significant and efficient predictor of perceived stress. In our study, the quantity of all types of green space, public and private, and the ways in which it can be used, appear to make a difference to community-level stress. Encouraging visits to local green space throughout the year also appears to be important for general health.