A cross-sectional analysis of green space prevalence and mental wellbeing in England

Mass migration and population growth over the last century have led to more than half of the world’s population residing in cities, creating a challenge for urban planners to efficiently accommodate new residents in a health promoting environment [1‐3]. It has been suggested that mental health may differ between urban and rural areas, with studies contrasting in the direction of their conclusions [4‐6]. Positive mental health and wellbeing have been linked to increased longevity, productivity and societal prosperity, but have also grown in prominence both politically and economically [7‐12]. For example, the EU-level Beyond GDP (Gross Domestic Product) initiative was developed to be more inclusive of such social and environmental aspects of progress, by quantifying climate change, poverty and mental wellbeing, as well as the economy [13]. In the UK, results from the 2015 Annual Population Survey showed that, while mental wellbeing had on average increased over recent years, the divide between those rating their personal wellbeing at the highest and lowest levels had also grown, indicating a wellbeing inequality which needs to be addressed [14].

Mental wellbeing comprises two main components: the hedonic dimension, which includes happiness, life satisfaction and pain avoidance; and the eudaimonic dimension, which focuses on self-realisation, purpose in life and psychological function [15, 16]. Rather than just the absence of mental illness, mental wellbeing therefore encompasses aspects of positive affect, relaxation, functioning, personal relationships, life satisfaction and general happiness [17‐19].

Emerging evidence suggests that aspects of the physical environment, and exposure to nature in particular, are often associated with higher levels of happiness and life satisfaction [20‐22]. While these are important aspects of mental wellbeing, the relationship between green space and this multi-dimensional view of mental wellbeing remains relatively unexplored [20‐24].

In urban environments, green space is considered to be any area of grass, trees or other vegetation, which in towns and cities is deliberately reserved for recreational, aesthetic or environmental purposes; this term therefore covers a range of green urban features, including parks, sports pitches and streetscape greenery. While abundant in rural areas, green spaces are usually designed into urban landscapes, typically at the expense of buildings. To encourage this to happen, the UK government sets out green space recommendations to encourage Councils to build these into each neighbourhood; these recommendations have been developed from government survey-based research and consideration of accepted walking distances between homes and green spaces [25].

Studies have sought to understand why green spaces seem to be beneficial for health and wellbeing. The theory of biophilia suggests that people pursue connections to nature; humans evolved in a natural landscape, where green spaces would have offered shelter, potential sources of food, and hence survival, so we may still experience positive feelings in such environments [26, 27]. Exposure to nature might enhance wellbeing by providing mental escape and restoration from fatigue, which is the focus for two key theories. Attention Restoration Theory proposes that effortful, directed attention is required to undertake everyday tasks, while the involuntary fascination which nature attracts provides an opportunity to rest the brain and regain concentration [28‐30]. By contrast, it is suggested that urban environments may be less restorative, because of excessive stimuli and a need for directed attention to process these high levels of information [29, 31]. An alternative, the Stress Recovery Theory, argues that views of nature are the most beneficial for restoration, by helping stressed individuals recover a relaxed emotional state [32, 33]; these theories have been validated by a number of studies [31, 34‐40]. It perhaps follows that individuals are often attracted to scenic environments, in particular trees, vegetation and water [1, 32, 41, 42], and so exposure to such landscapes may be valuable for happiness [22, 33, 43‐45]. As well as these restorative mechanisms, it is theorised that green spaces may contribute to better health by enabling activities known to promote mental wellbeing, such as social interaction [2, 19, 46, 47] and physical activity [21, 48].

Recent research has begun investigating the association between the proportion of green space in neighbourhoods and residents’ mental health and wellbeing [8, 21, 23, 49, 50]. One study found a positive association to a single life satisfaction measure, by analysing 10,000 individuals living in Lower-Layer Super Output Areas (LSOAs) in urban England [22]. Other work has demonstrated that socioeconomic inequalities in mental wellbeing (indexed by the WHO-5 positive wellbeing index) tend to be smaller among those who feel they have good access to recreational areas within their urban neighbourhood, although this study did not objectively quantify green space, or restrict recreational areas to those that were specifically green [8]. Several studies also report that people are more likely to have lower levels of mental distress, as measured by the General Health Questionnaire (a psychiatric screening tool), when residing in areas with relatively more green space [22, 23, 51]. One such longitudinal study reported that ward-level proportions of green space were negatively associated with psychiatric morbidity, although the strength of this association varied across life course and by gender [52]. While lower levels of psychiatric symptoms are generally associated with better wellbeing, as described, mental wellbeing is a positive measure which reflects much more than an absence of distress [53].

While studies in this area tend to examine aspects of positive mental health, such as relaxation, satisfaction and general happiness [1, 22, 32, 41‐45, 54, 55], we are only aware of one other study implementing a multi-dimensional measure of mental wellbeing. The study was based on a small selective sample in deprived areas of Scotland, and investigated the association between local green space proportions and mental wellbeing, of which the results were mixed and inconclusive [56].

Previous studies have tended to consider either urban green space or the wider benefits of contact with nature; while urban-rural differences in health have been studied, it is not yet known whether the association between green space and mental wellbeing in particular differs in urban and rural areas [1, 31, 32, 57, 58]. Although urbanisation reduces opportunities for people to interact with natural environments, it remains unclear whether or how this might affect the mental wellbeing of those who live in cities [59, 60].

The primary aim of this research was to test two hypotheses: (1) that neighbourhood areas of England with greater proportions of local-area green space are associated with higher levels of mental wellbeing; and (2) that the association between the proportion of local area green space and mental wellbeing may be confounded and/or modified by urban versus rural location.

In total, 50,994 individuals were included in wave 1 of the study, from 30,169 different households, which equates to a 57.6% participation response from the initially selected households, followed by an 81.8% individual-level response rate to the questionnaires issued to these agreeing households [72]. Little direct information was available regarding the characteristics of non-responding individuals, although they may be compared in terms of local-area socioeconomic statistics. The data collectors (Understanding Society) observed slightly lower response rates in areas with higher proportions of single-person households (59.0% response in 1st quartile of single-person households, compared to 55.5% in the highest,4th, quartile) and people in full-time employment (59.7% response in 1st quartile, 56.6% in 4th). Similarly, at the individual level, response rates were somewhat higher in areas of lower deprivation, in terms of Council Tax band (86.2% response in the lowest band A, 79.5% response in the highest bands E-H), suggesting a modest association between socio-economic status and survey participation [72].

Of the responding individuals, 42,972 were residents of England. After removing those who had missing SWEMWBS (mental wellbeing) scores, the final sample contained 30,900 individuals, from 19,684 different households, which is 61.0% of the original sample from the UKLHS. The sample covers 11,096 LSOAs across England, which vary considerably in size between urban (mean 0.9km², sd 2.3km²) and rural areas (mean 19.6km², sd 25.1km²). Of those not completing the mental wellbeing questions, mean green space exposure was 0.36 (sd 0.28), which was lower than the final sample (mean 0.42, sd 0.30) (Significance of t-test, p < 0.001).

From a socioeconomic perspective, local-area deprivation was significantly greater among SWEMWBS non-completers (mean score 27.1, sd 17.2 versus, 22.2, sd 15.6)(p < 0.001), although average equivalised income was consistent (£5515/month, sd £5438 for responders versus £5511/month, sd £5970 for non-responders) (p = 0.831).

The characteristics of people living in urban (n = 25, 547) and rural (n = 5353) areas also differed. The mean age of respondents was higher in rural areas, which also had greater proportions of married individuals. Income was also higher in rural areas, where area-level deprivation was considerably lower, household space was greater and more people owned their own home. These findings are presented in Table 1; t-tests were used to estimate the significance of the difference between urban are rural variables.

The unadjusted regression coefficient, B, for the association between proportion of green space and mental wellbeing was 0.17 points (95% CI 0.11, 0.23) in the SWEMWBS score, per standard deviation increase in green space. After controlling for all individual and household-level confounding factors (apart from urban/rural location), this coefficient was reduced 0.01 points (−0.05, 0.07) (p = 0.774).

Finally, adjusting further for urban/rural location in the association between a standard deviation increase in green space and SWEMWBS score, the resultant B value was −0.01 points (−0.08, 0.5, p = 0.712). While green space and urbanity were significantly linearly associated (B = −0.23, p < 0.001), we only found slight, but statistically insignificant evidence of effect modification (B = −0.11, 95% CI -0.29, 0.11, p = 0.382) between these variables. Stratified univariate models showed that the association was slightly stronger in rural (B = 0.12 points, 95% CI -0.01, 0.21, p = 0.062) than urban areas (B = 0.07 points, 95% CI 0.01, 0.13, p = 0.027), for a standard deviation increase in green space, although only the urban result was statistically significant.

The results of the fully-adjusted model are presented in Table 2.

As a sensitivity analysis, we repeated these models using quasi-poisson and log-transformed regressions, to account for the skewed distribution of the SWEMWBS variable. These modelling techniques did not significantly change our findings.

The proportion of green space in an individual’s local area was significantly and positively associated with mental wellbeing in univariate models, but became weaker and statistically non-significant after adjusting for socio-demographic variables and urban/rural location. While the green space in an individual’s local area has been shown to be related to aspects of mental health such as happiness and life satisfaction, the association to multi-dimensional mental wellbeing is much less clear. Further research is therefore needed to explore the relationship of other aspects of green spaces aside from size, such as accessibility, aesthetics and use, to mental wellbeing.