Content uploaded by Jo Garrett
Author content
All content in this area was uploaded by Jo Garrett on Oct 11, 2019
Content may be subject to copyright.
Contents lists available at ScienceDirect
Health & Place
journal homepage: www.elsevier.com/locate/healthplace
Coastal proximity and mental health among urban adults in England: The
moderating effect of household income
Joanne K. Garrett
a,∗
, Theodore J. Clitherow
a
, Mathew P. White
a
, Benedict W. Wheeler
a
,
Lora E. Fleming
a
a
European Centre for Environment and Human Health, European Centre for Environment and Human Health, University of Exeter Medical School, Knowledge Spa, Royal
Cornwall Hospital Treliske, Truro, Cornwall, TR1 3HD, UK
ABSTRACT
After adjusting for covariates, self-reported general health in England is higher among populations living closer to the coast, and the association is strongest amongst
more deprived groups. We explored whether similar findings were present for mental health using cross-sectional data for urban adults in the Health Survey for
England (2008–2012, N ≥25,963). For urban adults, living ≤1 km from the coast, in comparison to > 50 km, was associated with better mental health as measured
by the GHQ12. Stratification by household income revealed this was only amongst the lowest-earning households, and extended to ≤5 km. Our findings support the
contention that, for urban adults, coastal settings may help to reduce health inequalities in England.
1. Introduction
1.1. Overview
Poor mental health is among the leading causes of disability
worldwide (World Health Organisation, 2018). In England, approxi-
mately one in six adults (17 %) surveyed were suffering symptoms of a
common mental disorder (CMD), such as anxiety or depression
(McManus et al., 2016). However, there is mounting evidence that
exposure to natural environments is associated with various benefits for
mental health and wellbeing (hereby referred to as ‘mental health’; see
reviews by Bratman et al., 2012;Frumkin et al., 2017;Hartig et al.,
2014;Trostrup et al., 2019). Much of this work reports positive asso-
ciations between green space and mental health, where measured or
tested exposures include neighbourhood vegetation, green exercise, and
residential proximity to green space (e.g. Barton and Pretty, 2010;
Beyer et al., 2014;Cox et al., 2017b;de Vries et al., 2013;Gascon et al.,
2015;McEachan et al., 2016). This may be particularly the case for
those in urban areas where nature exposures can be limited (Cox et al.,
2017a).
Concurrently, a smaller, yet growing, amount of research suggests
that blue spaces (aquatic environments such as coasts, rivers, and lakes)
are associated with a range of aspects related to improved mental
health. These include: enhanced general health and wellbeing (re-
viewed by Gascon et al., 2017; see also Wheeler et al., 2012;White
et al., 2013a;Völker et al., 2018;Volker and Kistemann, 2011;Wood
et al., 2016); increased physical activity levels (White et al., 2014);
improved psychological restoration (White et al., 2010;White et al.,
2013b); reduced psychological distress (Nutsford et al., 2016); and
lower mortality rates (Crouse et al., 2018). A range of blue space ex-
posures have been explored in these studies including area coverage,
presence/absence, visibility and perceived and objective proximity.
There is also evidence that socioeconomic status may act as an ef-
fect-modifier, or moderator, of the nature-health relationship (see
Hartig et al., 2014;Markevych et al., 2017;Mitchell et al., 2015). For
example, several cross-sectional studies find that the association be-
tween natural environments and mental health is stronger within more
deprived areas, or that health inequality gradients are lessened where
green/blue space is more available (e.g. Wheeler et al., 2012;Maas
et al., 2006;McEachan et al., 2016;Mitchell and Popham, 2008;
Mitchell et al., 2015;van den Berg et al., 2016;Ward Thompson et al.,
2012; however, see also Mitchell and Popham, 2007).
Again, however, most of this work has examined socioeconomic
deprivation as a moderator of health regarding various measures of
green space, with exposure to blue spaces receiving less empirical in-
vestigation (Markevych et al., 2017). Indeed, to the best of the authors’
knowledge, only two studies have explicitly tested this relationship.
First, Wheeler et al. (2012) found that the relationship between living
closer to the coast in England and self-reported general health was
strongest amongst communities within areas of higher socioeconomic
deprivation. More recently, Crouse et al. (2018) examined the asso-
ciation between blue space and mortality in Canada, with results sug-
gesting a similar pattern of effect-modification but lacking statistical
power for some outcomes. A further study investigated the moderating
https://doi.org/10.1016/j.healthplace.2019.102200
Received 21 January 2019; Received in revised form 16 July 2019; Accepted 22 August 2019
∗
Corresponding author. ECEHH, University of Exeter Medical School, Knowledge Spa, Royal Cornwall Hospital Treliske, Truro, Cornwall, TR1 3HD, UK.
E-mail address: j.k.garrett@exeter.ac (J.K. Garrett).
Health and Place xxx (xxxx) xxxx
1353-8292/ © 2019 Published by Elsevier Ltd.
Please cite this article as: Joanne K. Garrett, et al., Health and Place, https://doi.org/10.1016/j.healthplace.2019.102200
effect of educational attainment, one aspect of socioeconomic status, on
the relationship between both blue spaces and green spaces on various
mental and physical health outcomes. They found a significant inter-
action between blue space and both health outcomes for those with the
lowest educational attainment (de Vries et al., 2016).
Thus, although there have been encouraging findings, research ex-
amining the links between blue space and mental health remains lim-
ited (Gascon et al., 2017). Furthermore, despite growing health in-
equalities (Barr et al., 2015;Thomson et al., 2018), we currently have a
poor understanding of how this relationship might vary between dif-
ferent levels of socioeconomic deprivation (Mitchell et al., 2015).
1.2. The current research
The aim of the present research was to investigate: (1) the asso-
ciation between mental health, as measured using two different in-
dicators of common mental disorders (CMD), and the blue space ex-
posure of coastal proximity, as used in Wheeler et al. (2012) for urban
residents; and (2) variations in this association according to household
income. The study therefore aimed to directly build on work by
Wheeler et al. (2012) through focusing on self-reported mental health
(as opposed to general health) as the dependent variable and at
household level (instead of area level) deprivation in the form of
household income, as the moderating variable. We used the Health
Survey for England (HSE), a comprehensive, nationally representative
survey which includes various measures of health, health behaviours
and socio-demographics (Aresu et al, 2009,2010,2011;Boniface et al.,
2012;Bridges et al., 2013). Based on the literature previously in-
troduced, we hypothesised that: (a) CMD likelihood would decrease as
coastal proximity increased; and (b) this association would be stronger
amongst lower income households.
2. Methods
2.1. Sample
Secondary cross-sectional data were utilised from the HSE for
English adults for the years 2008–2012 (pooled; aged 16+; adults
N = 45,063). All inhabitants of selected households are eligible for
interview and full sampling details can be found in (Aresu et al, 2009,
2010,2011;Boniface et al., 2012;Bridges et al., 2013). Trained inter-
viewers ask respondents a set of core questions related to their health,
lifestyle, and background, with additional sections which vary each
year.
The GHQ12 was not included in 2011 (n adults 2008–2010 and
2012 = 36,453), or the EQ5D in 2009 (n adults 2008 and
2010–2012 = 40,418), reducing the available samples for our analyses.
Both the prevalence of mental health disorders and the relationships
between natural environments and health have been found to vary by
urbanity (see Alcock et al., 2015;Maas et al., 2006;Mitchell and
Popham, 2007;Peen et al., 2010;Wheeler et al., 2012;Wood et al.,
2016). Further, access to health services (Bauer et al., 2018;Chukwusa
et al., 2019) and characteristics of natural environments are very dif-
ferent between urban and rural areas. We therefore focused only on
urban residents, which are those individuals categorised by the trained
interviewers as living in an ‘Urban’setting, as opposed to ‘Rural or
isolated dwellings’or ‘Town and Fringe’(Bibby and Brindley, 2013).
Available urban adults sample sizes were 28,662 (GHQ12) and 31,906
(EQ5D). For respective analyses, we excluded those with missing re-
sponses for GHQ12 and the anxious/depression dimension of the EQ5D,
therefore the full samples were 26,099 and 28,885 for GHQ12 and
anxious/depression respectively.
The richness of the dataset enabled the inclusion of a range of po-
tential confounding factors which may also relate to mental health,
including: income, age, sex and the presence of limiting longstanding
illnesses. We also included the health risk factors smoking status and
body mass index (BMI), as these have received limited attention in
previous studies exploring environment-health relationships (Mitchell,
2013).
We categorised responses of “Item not applicable”,“No answer/
refused”and “Don't know”as missing and calculated the sample sizes.
Missing data categories were excluded where there were < 20 re-
spondents in a category (see Supplemental Table 2). This led to final
analysis sample sizes of 25,963 (GHQ12) and 28,723 (EQ5D Anxiety/
depression).
2.2. Coastal proximity
Following previous approaches (Wheeler et al., 2012;White et al.,
2013a), coastal proximity was measured in terms of the Euclidean
distance (km) from the population density weighted centroid of re-
spondents' Lower-layer Super Output Area (LSOA, as at 2001 Census) to
the nearest coastline. There are approximately 32,500 LSOAs in Eng-
land, each with a mean area of 4 km
2
and containing an average po-
pulation of around 1500 (Wheeler et al., 2012). Following Wheeler
et al. (2012), we operationalised coastal proximity using five cate-
gories: (1) 0–1 km; (2) > 1–5 km; (3) > 5–20 km; (4) > 20–50 km;
(5) > 50 km. As with previous studies (Wheeler et al., 2012;White
et al., 2013a), we used > 50 km as the reference category to enable us
to test if the likelihood of having a CMD decreases with proximity to the
coast. This also allowed us to compare ‘coastal’(i.e. < 50 km; as used in
EU definitions e.g. defining coastal regions (Eurostat, 2018)) re-
spondents with ‘inland’(i.e. ≥50 km) respondents (White et al.,
2013b).
2.3. Self-reported mental health
Mental health was measured through two outcomes. The first was
the 12-item version of the General Health Questionnaire (GHQ12;
Goldberg et al., 1997), available 2008–2010 and 2012, a self-reported
measure widely used by health practitioners and researchers to indicate
the likelihood or ‘caseness’of an individual having a high risk of a CMD.
Following established recommendations for the GHQ12, results were
dichotomised with scores of four or above widely considered predictive
of a high risk of common mental health disorders such as anxiety or
depression (Fryers et al., 2004;Katikireddi et al., 2012;Mann et al.,
2011;Semlyen et al., 2016). The two outcome categories for this
measure were therefore: high likelihood of a CMD (GHQ12 score ≥4);
and low likelihood of a CMD (GHQ12 score < 4).
The second outcome was the anxiety and depression dimension of
the EQ-5D-3L (hereafter referred to as the EQ5D; EuroQol Research
Foundation, 2018). The EQ5D is a standardised measure of health-re-
lated quality of life (EuroQol Research Foundation, 2018), in-
corporating five dimensions, which has been used by practitioners and
researchers (EuroQol, 2018;Hulme et al., 2004;Park et al., 2011) and
utilised in studies exploring environmental characteristics (de Oliveira
et al., 2013;Kyttä et al., 2011). Although the intended use is as a
composite scale, here we use a single dimension –anxiety and de-
pression. This dimension has been associated with anxiety and/or de-
pression measured using a diagnostic scale (Mini International Neu-
ropsychiatric Interview (Supina et al., 2007)), found to align with the
GHQ12 (Bohnke and Croudace, 2016) and used for the same purpose in
other studies (Semlyen et al., 2016). It should be noted that it was not
found to be responsive to changes in anxiety or depression for those
clinically diagnosed (Crick et al., 2018) and therefore not necessarily a
measure of clinical diagnoses of anxiety and depression. As with the
GHQ12, it is a self-completed scale. There are three possible response
options, with respondents reporting whether they are not anxious or
depressed; moderately anxious or depressed; or extremely anxious or
depressed at the time of completion. These responses were dichot-
omised into the categories ‘Not anxious/depressed’and ‘At least mod-
erately anxious/depressed’(moderate or extreme anxiety/depression)
J.K. Garrett, et al. Health and Place xxx (xxxx) xxxx
2
to account for the skewness in data and low sample sizes within the
extremely anxious depressed category (n = 648; 2 % of total).
2.4. Area level controls
In line with previous research (Wheeler et al., 2012;White et al.,
2017;Mitchell and Popham, 2007), we controlled for area level de-
privation (English Index of Multiple Deprivation, Noble et al., 2007), as
well as green space and freshwater coverage at LSOA level to explore
the unique effect of coastal proximity. The English Index of Multiple
Deprivation (IMD) consists of area measures of crime, employment,
education, and income and has been found to be related to mental
health (Bellis et al., 2012) and moderate the coastal-health relationship
(Wheeler et al., 2012). Percentage greenspace coverage was based on
the generalised land use database (GLUD; Department for Communities
and Local Government, 2007) for LSOAs and incorporated all area level
green spaces, not including private gardens. Percentage freshwater
coverage of the LSOA was derived from the CEH Land Cover Map 2007
(Morton et al., 2011).
2.5. Household level controls
We also included household and individual level covariates which
may also relate to mental health. Equivalised household income, which
takes into account the number of household members, was used to
assess household level deprivation (reference category = highest in-
come quintile). Household income has been found to be related to a
range of mental health disorders (Domenech-Abella et al., 2018;Kahn
et al., 2000;Sareen et al., 2011). The upper and lower bounds of each
quintile vary by year and are given in Supplemental Table 1. Car access
was also included at the household level (ref = access).
2.6. Individual level controls
Individual level controls were based on confounders of mental
health identified by similar research with large survey datasets
(Wheeler et al., 2012;White et al., 2013a;White et al., 2013b;Beyer
et al., 2014;Crouse et al., 2017;Stranges et al., 2014). These included:
sex (reference = female), age (reference = 16–34 years old), highest
qualification level (reference = none/foreign/other), economic status
(reference = in employment/student), relationship status (re-
ference = single), year (reference = 2008), presence of limiting long-
standing illnesses (reference = no limiting longstanding illness), ci-
garette smoking status (reference = never smoked cigarettes at all), and
weight (body mass index; BMI; reference = normal weight).
2.7. Data linkage
Standard licence versions of HSE data only include large area geo-
graphical identifiers to preserve anonymity. In order to allocate higher
resolution measures of coastal proximity, green space and freshwater,
these three variables at LSOA level were supplied by the authors to the
data providers (NatCen Social Research) and linked anonymously to
HSE data under agreement from the NHS Health and Social Care
Information Centre (now NHS Digital). To prevent identification of any
individual LSOA of residence, the three environmental variables were
constrained to relatively coarse categories; and LSOA and regional
identifiers were removed from the linked data and returned to the au-
thors.
2.8. Analyses
Data were analysed using the “survey”package (version 3.34;
Lumley, 2018) in R Studio Version 3.4.2. Generalised linear models
(GLM) using a quasi-binomial error structure (appropriate when ana-
lysing complex survey data (Lumley, 2018)) and household clusters, to
account for multiple respondents within households and provide robust
standard errors, were used to identify correlations between coastal
proximity and mental health. We were not able to include clustering by
LSOA as this had been removed by the data providers for anonymity.
The data were weighted using the interview weights provided in the
dataset to account for selection, non-response and population biases
(Aresu et al, 2009,2010,2011;Boniface et al., 2012;Bridges et al.,
2013). We calculated the odds ratios (OR) and 95% confidence intervals
(CI) of participants having either a high likelihood of a CMD
(GHQ12 ≥4) or of reporting a status of at least moderately anxious/
depressed for this dimension of the EQ5D.
We present unadjusted models (nature exposures only) and fully-
adjusted models to examine how coastal proximity was associated with
mental health before and after adding the controls. A sensitivity ana-
lysis was also carried out with > 20 km as a reference category. We
then stratified our analysis by household income, whereby we analysed
the relationships between coastal proximity and mental health using
fully-adjusted (unweighted) GLMs for each household income quintile.
This enabled us to observe variations in the relationship between
coastal proximity and mental health by household income. We had an a
priori prediction that the effects would be strongest in the lowest in-
come quintiles, however, we also carry out analyses interacting coastal
proximity and household income.
3. Results
3.1. Full model/sample results
Table 1 presents descriptive statistics of the un-stratified mental
health models. In the GHQ12 model sample, the proportion of people
with a high risk of a common mental disorder (CMD) closely resembled
previous national averages (McManus et al., 2016), with approximately
15 % of participants reporting a high likelihood of suffering from a
CMD. In comparison, CMD prevalence was slightly higher in the EQ5D
model, likely due to the different method of measurement, with ap-
proximately 22 % of respondents reporting at least moderate anxiety or
depression. CMD prevalence was also greater amongst more deprived
areas and lower earning households (Table 1). For the years 2008, 2010
and 2012 where both the GHQ12 and anxiety and depression were
present, the correlation was 0.50 (kendall's τ,p< 0.001).
Table 2 displays the unadjusted and adjusted odds ratios (OR) with
95% confidence intervals (CI) of respondents having a high risk of CMD
for the full model samples (un-stratified) of both outcomes. Re-
spondents were less likely to report an at risk GHQ12 score of ≥4if
they lived up to 1 km of the coast compared to > 50km (OR
adj
= 0.78,
95 % CI = 0.65 –0.95).
No significant (p< 0.05) associations were found between coastal
proximity and CMD likelihood for either the GHQ12 outcome or an-
xiety/depression EQ5D dimension in the unadjusted models. Similarly,
there were no significant associations between coastal proximity and
the anxious/depression dimension of the EQ5D in the adjusted model.
Respondents living in areas of 80–100 % greenspace were less likely
to report at risk scores of the GHQ12 and being at least moderately
anxious or depressed in the unadjusted models (and 60–80 % with the
EQ5D dimension). However, neither of these associations held in the
adjusted models.
Freshwater coverage of > 5–100 % was found to be related to the
anxious/depression dimension of the EQ5D in both the unadjusted and
adjusted models (OR
adj
= 0.78, 95 % CI = 0.63 –0.96). However, it
was not related to GHQ12 in either the unadjusted or adjusted model.
The results from our sensitivity analysis with > 20 km as a reference
category were similar, giving us confidence in our results
(Supplementary Table 3). As with a reference category of > 50 km, we
find significant associations between living ≤1 km from the coast and
the GHQ12 (≤1km vs. >20km OR
adj
= 0.79, 95 % CI = 0.66–0.94).
J.K. Garrett, et al. Health and Place xxx (xxxx) xxxx
3
Table 1
Descriptive statistics for the un-stratified GHQ12 (N = 25,963) and EQ5D (N = 28,723) models.
Variables GHQ12 EQ5D
Full model sample Low risk of CMD
a
(score < 4) High risk of CMD
a
(score ≥4) Full model sample Not anxious or depressed At least moderately anxious or
depressed
Unweighted N Weighted %
age
Unweighted N Weighted %
age
Unweighted N Weighted %
age
Unweighted N Weighted %
age
Unweighted N Weighted %
age
Unweighted N Weighted %age
Total Ns 25963 21984 84.88 3979 15.12 28723 22275 78.38 6448 21.62
LSOA
b
level variables
Coastal proximity
0–1km 1532 5.75 1315 86.03 217 13.97 1826 6.17 1413 78.01 413 21.99
>1–5km 3202 11.68 2729 85.70 473 14.30 3394 11.27 2635 78.88 759 21.12
>5–20 km 3781 13.46 3186 84.77 595 15.23 4400 13.93 3372 77.70 1028 22.30
>20–50 km 7252 30.10 6171 85.07 1081 14.93 7921 29.71 6216 79.13 1705 20.87
> 50 km (ref) 10196 39.01 8583 84.36 1613 15.64 11182 38.92 8639 77.97 2543 22.03
Freshwater coverage
>5–100 % 712 2.74 602 84.83 110 15.17 776 2.67 627 81.89 149 18.11
>1–5% 1607 6.09 1386 86.81 221 13.19 1790 6.10 1393 78.19 397 21.81
>0–1% 1459 5.42 1244 85.38 215 14.62 1658 5.63 1278 77.81 380 22.19
0 % (ref) 22185 85.75 18752 84.72 3433 15.28 24499 85.59 18977 78.32 5522 21.68
Greenspace coverage
80–100 % 1697 6.03 1478 87.33 219 12.67 1897 6.15 1516 80.20 381 19.80
60 - < 80 % 3239 11.84 2788 86.23 451 13.77 3539 11.76 2817 80.21 722 19.79
40 - < 60 % 4570 17.04 3842 84.41 728 15.59 5108 17.15 3912 77.26 1196 22.74
20 - < 40 % 7630 29.15 6398 84.11 1232 15.89 8545 29.45 6547 77.97 1998 22.03
0 - < 20 % (ref) 8827 35.93 7478 84.88 1349 15.12 9634 35.49 7483 78.34 2151 21.66
IMD
c
Most deprived 5549 21.65 4410 80.26 1139 19.74 6219 21.97 4495 73.83 1724 26.17
2nd most deprived 5540 21.90 4640 84.23 900 15.77 6133 21.82 4674 77.70 1459 22.30
Medium deprived 5051 19.79 4298 85.01 753 14.99 5566 19.67 4343 78.60 1223 21.40
2nd least deprived 4573 17.24 3986 87.30 587 12.70 5118 17.53 4086 80.53 1032 19.47
Least deprived (ref) 5250 19.43 4650 88.49 600 11.51 5687 19.00 4677 82.21 1010 17.79
Household level variables
Household income quintile
Lowest 3922 14.76 2969 76.12 953 23.88 4347 14.86 2890 68.07 1457 31.93
Second lowest 4172 15.14 3465 83.27 707 16.73 4662 15.37 3501 76.05 1161 23.95
Middle 4167 15.74 3576 86.22 591 13.78 4640 15.73 3677 80.06 963 19.94
Second highest 4434 17.48 3917 88.18 517 11.82 4853 17.27 3982 82.15 871 17.85
Missing data 4923 19.76 4159 84.38 764 15.62 5478 19.97 4191 77.40 1287 22.60
Highest (ref) 4345 17.12 3898 89.85 447 10.15 4743 16.80 4034 85.34 709 14.66
Car access
No 5503 20.79 4277 78.04 1226 21.96 6165 20.91 4173 68.92 1992 31.08
Yes (ref) 20460 79.21 17707 86.68 2753 13.32 22558 79.09 18102 80.88 4456 19.12
Individual level variables
Age categories
75+ 2615 8.13 2188 83.63 427 16.37 2875 8.01 2110 73.24 765 26.76
55 - 74 7344 23.65 6328 86.01 1016 13.99 8036 23.56 6132 76.47 1904 23.53
35 - 54 9049 35.18 7558 83.93 1491 16.07 10060 35.25 7702 77.15 2358 22.85
16–34 (ref) 6955 33.05 5910 85.40 1045 14.60 7752 33.18 6331 82.28 1421 17.72
Highest qualification
Higher ed/Degree 8203 32.73 7130 86.83 1073 13.17 9294 33.27 7596 81.89 1698 18.11
NVQ3/A level 3944 16.98 3364 85.64 580 14.36 4429 17.19 3609 82.27 820 17.73
NVQ1/NVQ2/GCSE 6989 26.97 5888 84.66 1101 15.34 7650 26.81 5858 77.46 1792 22.54
Other/none (ref) 6827 23.33 5602 81.86 1225 18.14 7350 22.73 5212 71.38 2138 28.62
Working status
(continued on next page)
J.K. Garrett, et al. Health and Place xxx (xxxx) xxxx
4
Table 1 (continued)
Variables GHQ12 EQ5D
Full model sample Low risk of CMD
a
(score < 4) High risk of CMD
a
(score ≥4) Full model sample Not anxious or depressed At least moderately anxious or
depressed
Unweighted N Weighted %
age
Unweighted N Weighted %
age
Unweighted N Weighted %
age
Unweighted N Weighted %
age
Unweighted N Weighted %
age
Unweighted N Weighted %age
ILO unemployed
d
803 3.54 576 72.26 227 27.74 923 3.58 601 66.91 322 33.09
Retired/other inactive 9532 30.82 7634 79.43 1898 20.57 10430 30.64 7248 68.96 3182 31.04
In work/student (ref) 15628 65.64 13774 88.12 1854 11.88 17370 65.77 14426 83.40 2944 16.60
Sex
Male 11497 48.75 9986 86.90 1511 13.10 12668 48.76 10275 81.54 2393 18.46
Female (ref) 14466 51.25 11998 82.96 2468 17.04 16055 51.24 12000 75.37 4055 24.63
Relationship status
In a relationship 16207 61.27 14132 87.25 2075 12.75 17870 61.25 14427 81.06 3443 18.94
Widow/separated./
divorced
4484 14.60 3530 78.33 954 21.67 4980 14.58 3394 67.89 1586 32.11
Single (ref) 5272 24.13 4322 82.84 950 17.16 5873 24.17 4454 77.91 1419 22.09
Limiting illness presence
Limiting illness 22.13 4341 67.77 2024 32.23 7046 22.09 3981 56.64 3065 43.36
Non-limiting longstanding illness 18.22 4498 89.38 514 10.62 5442 17.95 4355 79.86 1087 20.14
No longstanding illness (ref) 59.65 13145 89.86 1441 10.14 16235 59.95 13939 85.95 2296 14.05
BMI-
e
Obese 5869 21.71 4837 82.65 1032 17.35 6517 21.87 4824 75.10 1693 24.90
Overweight 8412 31.93 7316 86.95 1096 13.05 9237 31.65 7328 79.84 1909 20.16
Underweight 373 1.70 294 80.26 79 19.74 399 1.63 298 77.55 101 22.45
Missing data 3386 12.37 2736 81.61 650 18.39 3835 12.76 2846 75.17 989 24.83
Normal weight (ref) 7923 32.30 6801 85.84 1122 14.16 8735 32.10 6979 80.49 1756 19.51
Smoking status
Current smoker 5587 22.19 4398 79.16 1189 20.84 6151 22.02 4279 70.53 1872 29.47
Used to smoke 8099 29.21 6915 85.46 1184 14.54 8952 29.20 6981 78.68 1971 21.32
Never smoked (ref) 12277 48.61 10671 87.15 1606 12.85 13620 48.79 11015 81.74 2605 18.26
Year
2012 5591 21.53 4702 84.31 889 15.69 5696 19.92 4500 79.80 1196 20.20
2011 - –––––5891 20.35 4277 73.78 1614 26.22
2010 5857 22.80 4956 84.71 901 15.29 5865 20.76 4473 76.94 1392 23.06
2009 3308 12.99 2728 82.68 580 17.32 ––––– –
2008 (ref 11207 42.68 9598 85.94 1609 14.06 11271 38.96 9025 80.82 2246 19.18
a
CMD = common mental disorder.
b
LSOA = Lower-layer Super Output Area;
c
IMD = Indices of Multiple Deprivation;
d
ILO = International Labour Organisation;
e
BMI = Body Mass Index.
J.K. Garrett, et al. Health and Place xxx (xxxx) xxxx
5
Table 2
Unadjusted and adjusted regression models predicting the likelihood of respondents having poor metal health as assessed using the GHQ12 and anxiety/depression
component of the EQ5D. Significant results are highlighted in bold type.
GHQ12 (≥4) EQ5D: at least moderately anxious/depressed
Unadjusted Adjusted Unadjusted Adjusted
term OR 95 % CI pOR 95 % CI pOR 95 % CI pOR 95 % CI p
LSOA level variables
a
Coastal proximity
0–1km 0.87 0.73–1.04 0.121 0.78 0.65–0.95 0.011 0.99 0.85–1.14 0.836 0.90 0.78–1.05 0.195
>1–5km 0.90 0.80–1.02 0.091 0.90 0.79–1.02 0.108 0.95 0.85–1.05 0.330 0.93 0.83–1.04 0.188
>5–20 km 0.97 0.86–1.09 0.592 0.99 0.87–1.12 0.845 1.02 0.93–1.12 0.709 1.01 0.92–1.11 0.818
>20–50 km 0.94 0.86–1.04 0.229 0.97 0.88–1.07 0.507 0.93 0.86–1.01 0.083 0.97 0.89–1.05 0.451
> 50 km (ref)
Freshwater coverage
>5–100 % 0.99 0.78–1.26 0.949 0.95 0.74–1.21 0.656 0.79 0.65–0.97 0.022 0.78 0.63–0.96 0.020
>1–5% 0.87 0.74–1.03 0.098 0.85 0.72–1.02 0.074 1.03 0.89–1.19 0.709 1.05 0.91–1.21 0.541
>0–1% 1.03 0.86–1.24 0.758 1.03 0.85–1.26 0.768 1.09 0.95–1.25 0.239 1.12 0.96–1.31 0.141
0 % (ref)
Greenspace coverage
80–100 % 0.82 0.69–0.97 0.023 0.90 0.75–1.08 0.267 0.86 0.75–0.99 0.043 0.87 0.75–1.02 0.082
60 - < 80 % 0.90 0.79–1.02 0.103 0.97 0.85–1.11 0.643 0.88 0.79 –0.98 0.019 0.91 0.82–1.02 0.122
40 - < 60 % 1.04 0.93–1.16 0.504 1.02 0.91–1.15 0.720 1.06 0.97–1.16 0.218 1.02 0.92–1.12 0.763
20 - < 40 % 1.06 0.96–1.16 0.263 1.05 0.95–1.16 0.387 1.01 0.94–1.10 0.741 0.99 0.91–1.07 0.781
0 - < 20 % (ref)
IMD
b
Most deprived 1.21 1.05–1.38 0.008 1.04 0.92–1.17 0.518
2nd most deprived 1.08 0.94–1.24 0.257 0.99 0.89–1.11 0.922
Medium deprived 1.17 1.03–1.34 0.020 1.07 0.96–1.20 0.230
2nd least deprived 1.05 0.91–1.21 0.509 1.00 0.90–1.13 0.939
Least deprived (ref)
Household level variables
Household income quintile
Lowest 1.40 1.19–1.64 < 0.001 1.37 1.19–1.56 < 0.001
Second lowest 1.24 1.06–1.44 0.007 1.19 1.05–1.36 0.007
Middle 1.15 0.99–1.33 0.074 1.12 0.99–1.27 0.082
Second highest 1.13 0.97–1.31 0.119 1.20 1.06–1.36 0.004
Missing data 1.21 1.04–1.40 0.012 1.22 1.08–1.39 0.002
Highest (ref)
Car access
No 1.15 1.03–1.27 0.010 1.24 1.14–1.36 < 0.001
Yes (ref)
Individual level variables
Age categories
75+ 0.41 0.34–0.50 < 0.001 0.55 0.46–0.65 < 0.001
55 - 74 0.54 0.47–0.63 < 0.001 0.78 0.69–0.88 < 0.001
35 - 54 1.00 0.89–1.12 0.984 1.23 1.12–1.35 < 0.001
16–34 (ref)
Highest qualification
Higher ed/Degree 1.19 1.05–1.34 0.005 0.97 0.88–1.08 0.601
NVQ3/A level 1.12 0.98–1.29 0.105 0.85 0.76–0.96 0.007
NVQ1/NVQ2/GCSE 1.03 0.92–1.15 0.649 0.95 0.86–1.04 0.251
Other/none (ref)
Working status
ILO unemployed
c
1.97 1.62–2.39 < 0.001 1.72 1.46–2.03 < 0.001
Retired/other inactive 1.47 1.32–1.64 <0.001 1.55 1.42–1.70 < 0.001
In work/student (ref)
Sex
Male 0.79 0.73–0.86 < 0.001 0.75 0.70–0.80 < 0.001
Female (ref)
Relationship status
In a relationship 0.80 0.71–0.89 < 0.001 0.78 0.71–0.86 < 0.001
Widow/separated./divorced 1.12 0.98–1.28 0.098 1.07 0.95–1.20 0.282
Single (ref)
Limiting illness presence
Limiting illness 4.28 3.90–4.69 < 0.001 4.15 3.84–3.50 < 0.001
Non-limiting longstanding illness 1.19 1.06–1.34 0.003 1.62 1.48–1.77 <0.001
No longstanding illness (ref)
BMI
d
Obese 1.08 0.97–1.21 0.140 1.07 0.98–1.17 0.154
Overweight 0.95 0.86–1.06 0.373 1.03 0.94–1.12 0.511
Underweight 1.23 0.92–1.65 0.161 1.04 0.79–1.36 0.794
Missing data 1.12 0.99–1.26 0.071 1.03 0.92–1.15 0.608
Normal weight (ref)
Smoking status
Current smoker 1.40 1.26–1.54 < 0.001 1.55 1.43–1.69 <0.001
(continued on next page)
J.K. Garrett, et al. Health and Place xxx (xxxx) xxxx
6
3.2. Results stratified by household income
We find some significant interactions between coastal proximity and
household income for both the GHQ12 measure and the anxiety/de-
pression of the EQ5D (Supplemental Table 4).
Full results for each income quintile are presented in supplementary
materials (Supplemental Tables 5–9), with a summary of the key coastal
proximity findings in Fig. 1. As can be seen, living near the coast
(≤5 km) is associated with lower ORs (than living > 50 km) of poor
mental health as measured by both the GHQ12 (0–1km OR
adj
= 0.58,
95 % CI = 0.39 –0.87; > 1–5km OR
adj
= 0.76, 95 % CI = 0.59 –0.98)
and the anxiety/depression sub-scale of the EQ5D (0–1km
Table 2 (continued)
GHQ12 (≥4) EQ5D: at least moderately anxious/depressed
Unadjusted Adjusted Unadjusted Adjusted
term OR 95 % CI pOR 95 % CI pOR 95 % CI pOR 95 % CI p
Used to smoke 1.13 1.03–1.24 0.011 1.09 1.01–1.17 0.033
Never smoked (ref)
Year
2012 1.17 1.06–1.30 0.003 1.12 1.02–1.23 0.022
2011 1.62 1.48–1.77 < 0.001
2010 1.11 1.00–1.23 0.053 1.32 1.21–1.45 < 0.001
2009 1.30 1.15–1.47 < 0.001
2008 (ref)
Intercept −1.68 −2.52 −1.25 −2.16
N 25963 25963 28723 28723
Households 16592 16592 18419 18419
AIC
e
22056.47 19951.43 29987.95 26890.79
Cox & Snell pseudo-R
2
(%) 0.1 8.1 0.1 10.5
a
LSOA = Lower-layer Super Output Area; bIMD = Indices of Multiple Deprivation; cILO = International Labour Organisation; dBMI = Body Mass Index;
eAIC = Akaike's Information Criterion.
Fig. 1. The relationship between coastal proximity (reference category > 50 km) and mental health for each household income quintile. Note: results are fully
adjusted; CMD likelihood presented as odds ratios with 95% confidence intervals. Full model results in Supplemental Tables 5–9.
J.K. Garrett, et al. Health and Place xxx (xxxx) xxxx
7
OR
adj
= 0.72, 95 % CI = 0.53 –0.99; > 1–5km OR
adj
= 0.78, 95 %
CI = 0.62 –0.99) for individuals in the lowest household income
quintile only. There were no other significant associations between
coastal proximity and mental health for those in the higher household
income quintiles.
4. Discussion
In sum, we have explored the association between two measures of
mental health and coastal proximity for urban English adults using four
years of pooled data from the Health Survey for England. After ad-
justing for a range of relevant covariates, those living 0–1 km from the
coast had significantly lower odds of being at high risk of a CMD, as
measured by the GHQ12 and compared to those living further than
50 km. Coastal proximity was not found to be related to the anxiety/
depression EQ5D dimension.
As predicted, income quintile was a strong predictor of mental
health outcomes, and other socioeconomic factors (e.g. employment,
relationship and smoking status) were also largely consistent with
earlier work (Katikireddi et al., 2016;Stranges et al., 2014). However,
we find BMI not to be significantly related to mental health contrasting
with research by Stranges et al. (2014) using the HSE. Conversely, we
find that those who used to smoke were more likely to have poorer
mental health whereas this was not found by Stranges et al. (2014).We
also find not having access to a car was significantly related with worse
mental health whilst this was not found in earlier HSE years (Riva et al.,
2011).
Stratifying by household income revealed that the relationship be-
tween coastal proximity and mental health outcomes was present only
for those with the lowest household incomes and extended to < 5 km.
Specifically, the results imply that people living in urban areas in the
lowest household income quintile are less likely to suffer from a
common mental disorder (CMD) such as anxiety or depression if they
live within 5 km of the coast, compared to those living in urban areas
further inland (> 50 km). In particular, living within 1 km of the coast
is associated with the strongest reductions in CMD likelihood for people
from the most economically deprived households. Respondents from
this category reported symptoms consistent with a CMD according to
the GHQ12 measure with odds that were 40 % less than those living
further than 50 km. This is a greater reduction in comparison to being
in a relationship (vs. single OR
adj
= 0.78, 95 % CI = 0.63–0.98).
These findings add to the growing evidence base linking blue
spaces, particularly coastal environments, with better health and
wellbeing (White et al., 2010;Wheeler et al., 2012;White et al., 2013a;
Crouse et al., 2018;Gascon et al, 2015;Gascon et al., 2017;Nutsford
et al., 2016;Volker and Kistemann, 2011). This study also highlights
the potentially beneficial link between coastal proximity and common
mental disorders, which have been highlighted as growing issues in
countries such as England (McManus et al., 2016). Given that increas-
ingly many people live by and visit the coast in many countries, and
even more of them reside in cities, such research is vital for environ-
mental and social policy (Elliott et al., 2018;Pelling and Blackburn,
2014).
This research also supports previous work which suggests that the
positive relationship between living in more natural environments and
mental health is stronger within more socioeconomically deprived
groups (e.g. Wheeler et al., 2012;Maas et al., 2009;Maas et al., 2006;
McEachan et al., 2016;Mitchell and Popham, 2008). It also extends
prior research that investigated the interaction with area level depri-
vation (Wheeler et al., 2012), by demonstrating that household income
moderates the association between coastal proximity and health, in this
case specifically mental health. This suggests that access to the natural
environment may, at least partly, offset the adverse health and well-
being outcomes associated with low incomes. Indeed, recent work by
Elliott et al. (2018) finds that recreational visits to the English coast,
particularly walking, are more likely to be made by people from some
lower socioeconomic backgrounds as compared to other natural en-
vironments. Subsequently, ensuring coastal environments are accessible
to more socioeconomically deprived communities could therefore help
to reduce health inequalities (Elliott et al., 2018).
Although not established in this study, it is plausible that there is a
causal relationship between coastal living and mental health. Indeed, it
could be that exposure to coastal environments improves mental health
through a range of potential mechanisms in the same way as has been
proposed for green space, such as through reduced stress, improved air
quality and immune functioning, and increased opportunities for social
contact and physical activity (Hartig et al., 2014;Markevych et al.,
2017). In support of this, de Bell et al. (2017) sought to test whether the
same mechanisms that have been proposed to explain the relationship
between green space and health also applied in blue space visits. Most
people identified psychological benefits or social interactions as the
most important perceived benefit from their most recent blue space
visit. Similarly, higher levels of blue space visibility were associated
with lower levels of psychological distress in Wellington, New Zealand,
whilst green space visibility was not found to be related (Nutsford et al.,
2016) and, in Ireland, a sea view was found to be related to lower
depression scores (Dempsey et al., 2018). Earlier work by (Bauman
et al., 1999) also suggests that living by the coast is associated with
increased opportunities for physical activity.
More recently, White et al. (2014) found that people in England who
lived closer to the coast were more likely to visit the coast and, sub-
sequently, achieve their recommended weekly physical activity levels.
Combined with the finding that approximately 271 million recreational
visits are made each year to coastal environments in England (Elliott
et al., 2018), this suggests that the mental health of English coastal
urban dwellers (who are more likely to visit the coast) is better than
those in urban areas inland because of certain salutary mechanisms,
such as physical activity.
There were several unexpected findings in our research. For in-
stance, in contrast to previous research (e.g. de Bell et al., 2017;
MacKerron and Mourato, 2013;Völker et al., 2018), we found that
whilst living in closer proximity to coastal environments was sig-
nificantly linked to improved mental health outcomes, living in areas
with more freshwater coverage was not related overall according to the
GHQ12 measure. However, freshwater coverage was related to the
anxiety/depression EQ5D dimension. Freshwater coverage may be
specifically related to anxiety and depression, while the GHQ12 mea-
sure is slightly broader (Jackson, 2007). Further, we found more people
were at least moderately anxious or depressed under the EQ5D measure
than were at high risk of a CMD as measured under the GHQ12, sug-
gesting that this measure is perhaps a more sensitive measure of mental
health.
Similarly, green space coverage was not consistently related to
mental health after adjusting for confounders, as with Nutsford et al.
(2016). This is despite growing evidence that living within greener
environments is positively connected to general mental health and
wellbeing (see Hartig et al., 2014), as well as more specific factors as-
sociated with mental health, such as reduced stress levels (Cox et al.,
2017c;Van den Berg et al., 2010), reduced rates of antidepressant
prescriptions (Taylor et al., 2015), and increased psychological re-
storation (White et al., 2013b).
As previously discussed, it may therefore be that coastal environ-
ments are particularly important for mental health in comparison to
green spaces. Similar conclusions were drawn in Hong Kong, where
blue space visits were linked to mental health whilst visits to green
spaces were not related (Garrett et al., 2019). Our result may also be
due to the coarse measures used here to assess green space coverage.
While the GLUD data are based on a high resolution cartographic da-
tabase, it does not capture any measures of quality or accessibility that
may be important modifiers of any health benefits of proximity to green
space (Wheeler et al., 2015;Markevych et al., 2017). However, a re-
lationship with self-reported health has previously been detected using
J.K. Garrett, et al. Health and Place xxx (xxxx) xxxx
8
a similar measure elsewhere (Mitchell and Popham, 2007).
4.1. Limitations and future work
Beyond the potential limitations associated with self-reporting
health (e.g. Lee and Dugan, 2015), the cross-sectional nature of our
study means the results should be interpreted cautiously before making
generalisations about a causal relationship between coastal proximity
and common mental disorders (Gascon et al., 2017). Future work
should therefore examine the potential factors mediating this link, such
as physical activity. More longitudinal and experimental research (e.g.
White et al., 2013a;White et al., 2015;Annerstedt et al., 2012) is also
needed to elucidate a causal relationship and determine whether living
by the coast for an extended period remains beneficial for mental
health, as well as if these coastal benefits are consistently greater than
living in areas with more green space and freshwater coverage. Further,
our measure does not capture variations in accessibility and quality
which can relate to visit frequency and mental health benefits (Garrett
et al., 2019;Wyles et al, 2016,2017).
We were also not able to account for clustering at the LSOA level
which may have resulted in smaller standard errors as we cannot ac-
count for some potential non-independence within the data. However,
we have included LSOA level controls including additional nature ex-
posures and IMD.
4.2. Conclusion
To summarise, we found that the relationship between coastal
proximity and mental health was strongest for those urban adults in
more deprived households. This builds on previous research in-
vestigating coastal proximity and health inequalities at the community
level. Our results therefore add further evidence that the coast might
act as a mental health resource, particularly for people living in more
socioeconomically deprived circumstances. Ensuring access to these
environments may therefore have a role to play in reduction of health
inequalities (Allen and Balfour, 2014). At a time of increasing urbani-
sation, mental health disorders and degradation of coastal and marine
environments, such research should be developed and translated to
inform relevant environmental, planning and public health policies.
Acknowledgements
The authors would like to thank Dr Lewis Elliott for his advice and
guidance with the data analyses. The authors also thank Laura Brown at
ScotCen Social Research for support with HSE data linkage and ap-
provals from HSCIC/NHS Digital (Data Sharing Agreement NIC-09479-
J9Z4G). HSE data are copyright © 2013, re-used with the permission of
The Health and Social Care Information Centre [now NHS Digital], all
rights reserved. This work was supported by the National Institute for
Health Research Health Protection Research Unit (NIHR HPRU) in
Environmental Change and Health at the London School of Hygiene and
Tropical Medicine in partnership with Public Health England (PHE),
and in collaboration with the University of Exeter, University College
London, and the Met Office. The funders had no role in the study de-
sign, analysis, interpretation of data, or decision to submit the article
for publication. The views expressed are those of the author(s) and not
necessarily those of the NHS, the NIHR, the Department of Health, or
Public Health England. JG's time on the manuscript was undertaken as
part of the BlueHealth project, which received funding from the
European Union’s Horizon 2020 research and innovation programme
[grant agreement No.: 666773]. Analyses and interpretation are solely
the responsibility of the authors, and not the funders or data providers.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.healthplace.2019.102200.
References
Alcock, I., White, M.P., Lovell, R., Higgins, S.L., Osborne, N.J., Husk, K., Wheeler, B.W.,
2015. What accounts for 'England's green and pleasant land'? A panel data analysis of
mental health and land cover types in rural England. Landsc. Urban Plan. 142, 38–46.
Allen, J., Balfour, R., 2014. Natural Solutions for Tackling Health Inequalities, vol. 55
University College London Institute of Health Equity.
Annerstedt, M., Ostergren, P.O., Bjork, J., Grahn, P., Skarback, E., Wahrborg, P., 2012.
Green qualities in the neighbourhood and mental health - results from a longitudinal
cohort study in Southern Sweden. BMC Public Health 12, 12.
Aresu, M., Bécares, L., Brage, S., Chaudhury, M., Doyle-Francis, M., Esliger, D., Fuller, E.,
Gunning, N., Hall, J., Hirani, V., Jotangia, D., Mindell, J., Moody, A., Ogunbadejo, T.,
Pickip, D., Reilly, N., Robinson, C., Roth, M., Wardle, H., 2009. Health survey for
England 2008, volume 2: methods and documentation. In: Craig, R., Mindell, J.,
Hirani, V. (Eds.), The NHS Information Centre for Health and Social Care, pp. 203.
https://files.digital.nhs.uk/publicationimport/pub00xxx/pub00430/heal-surv-phys-
acti-fitn-eng-2008-rep-v3.pdf.
Aresu, M., Boodhna, G., Bryson, A., Bridges, S., Chaudhury, M., Craig, R., Fuller, E.,
Green, F., Hall, J., Hirani, V., Jarvis, D., Johnson, A.M., Mercer, C., Mindell, J.,
Nardone, A., Robinson, C., Roderick, P., Roth, M., Stocks, J., Thompson, J., 2011. In:
R., C., Mindell, J. (Eds.), Health Survey for England 2010: Volume 2. Methods and
Documentation, pp. 256. https://files.digital.nhs.uk/publicationimport/pub03xxx/
pub03023/heal-surv-eng-2010-resp-heal-vol2-meth-rep.pdf.
Aresu, M., Chaudhury, M., Diment, E., Fuller, E., Gordon-Dseagu, V., Gunning, N.,
Mindell, J., Nicholson, S., Ogunbadejo, T., Robinson, C., Roderick, P., Roth, M.,
Shelton, N., Tabaassum, F., Wardle, H., 2010. Health survey for England - 2009,
health and lifestyles: volume 2, methods and documentation. In: Craig, R., Hirani, V.
(Eds.), Joint Health Surveys Unit, National Centre for Social Research, UCL, pp. 182.
https://files.digital.nhs.uk/publicationimport/pub00xxx/pub00414/heal-surv-heal-
life-eng-2009-rep-v3.pdf.
Barr, B., Kinderman, P., Whitehead, M., 2015. Trends in mental health inequalities in
England during a period of recession, austerity and welfare reform 2004 to 2013. Soc.
Sci. Med. 147, 324–331.
Barton, J., Pretty, J., 2010. What is the best dose of nature and green exercise for im-
proving mental health? A multi-study analysis. Environ. Sci. Technol. 44, 3947–3955.
Bauer, J., Muller, R., Bruggmann, D., Groneberg, D.A., 2018. Spatial accessibility of
primary care in England: a cross-sectional study using a floating catchment area
method. Health Serv. Res. 53, 1957–1978.
Bauman, A., Smith, B., Stoker, L., Bellew, B., Booth, M., 1999. Geographical influences
upon physical activity participation: evidence of a ‘coastal effect’. Aust. N. Z. J. Public
Health 23, 322–324.
Bellis, M.A., Lowey, H., Hughes, K., Deacon, L., Stansfield, J., Perkins, C., 2012.
Variations in risk and protective factors for life satisfaction and mental wellbeing
with deprivation: a cross-sectional study. BMC Public Health 12, 492.
Beyer, K., Kaltenbach, A., Szabo, A., Bogar, S., Nieto, F., Malecki, K., 2014. Exposure to
neighborhood green space and mental health: evidence from the survey of the health
of Wisconsin. Int. J. Environ. Res. Public Health 11, 3453–3472.
Bibby, P., Brindley, P., 2013. Urban and Rural Area Definitions for Policy Purposes in
England and Wales: Methodology. Fareham: Office for National Statistics.
Bohnke, J.R., Croudace, T.J., 2016. Calibrating well-being, quality of life and common
mental disorder items: psychometric epidemiology in public mental health research.
Br. J. Psychiatry 209, 162–168.
Boniface, S., Bridges, S., Craig, R., Darton, R., Fuller, E., Hancock, R., Henderson, C.,
Knott, C., Mandalia, D., Mindell, J., Moody, A., Morciano, M., Ng Fat, L., Oyebode,
O., Robinson, C., Sadler, K., Sutton, R., Wittenberg, R., 2012. Health survey for
England - 2011: methods and documentation. In: Craig, R., Mindell, J. (Eds.), Joint
Health Surveys Unit, National Centre for Social Research, UCL, pp. 212. https://files.
digital.nhs.uk/publicationimport/pub09xxx/pub09300/hse2011-methods-and-docs.
pdf.
Bratman, G.N., Hamilton, J.P., Daily, G.C., 2012. The impacts of nature experience on
human cognitive function and mental health. Ann. N. Y. Acad. Sci. 1249, 118–136.
Bridges, S., Doyle, M., Fuller, E., Knott, C., Mindell, J., Moody, A., Ryley, A., Scholes, S.,
Seabury, C., Wardle, H., Whalley, R., 2013. In: R., C., Mindell, J. (Eds.), Health
Survey for England 2012: Volume 2 Methods and Documentation, pp. 240. https://
files.digital.nhs.uk/publicationimport/pub13xxx/pub13218/hse2012-methods-and-
docs.pdf.
Chukwusa, E., Verne, J., Polato, G., Taylor, R., Higginson, I.J., Gao, W., 2019. Urban and
rural differences in geographical accessibility to inpatient palliative and end-of-life
(PEoLC) facilities and place of death: a national population-based study in England,
UK. Int. J. Health Geogr. 18, 11.
Cox, D.T., Hudson, H.L., Shanahan, D.F., Fuller, R.A., Gaston, K.J., 2017a. The rarity of
direct experiences of nature in an urban population. Landsc. Urban Plan. 160, 79–84.
Cox, D.T., Shanahan, D.F., Hudson, H.L., Fuller, R.A., Anderson, K., Hancock, S., Gaston,
K.J., 2017b. Doses of nearby nature simultaneously associated with multiple health
benefits. Int. J. Environ. Res. Public Health 14, 172.
Cox, D.T.C., Shanahan, D.F., Hudson, H.L., Plummer, K.E., Siriwardena, G.M., Fuller,
R.A., Anderson, K., Hancock, S., Gaston, K.J., 2017c. Doses of neighborhood nature:
the benefits for mental health of living with nature. Bioscience 67, 147–155.
Crick, K., Al Sayah, F., Ohinmaa, A., Johnson, J.A., 2018. Responsiveness of the anxiety/
depression dimension of the 3- and 5-level versions of the EQ-5D in assessing mental
health. Qual. Life Res. 27, 1625–1633.
Crouse, D.L., Balram, A., Hystad, P., Pinault, L., van den Bosch, M., Chen, H., Rainham,
D., Thomson, E.M., Close, C.H., van Donkelaar, A., Martin, R.V., Menard, R.,
J.K. Garrett, et al. Health and Place xxx (xxxx) xxxx
9
Robichaud, A., Villeneuve, P.J., 2018. Associations between living near water and
risk of mortality among urban Canadians. Environ. Health Perspect. 126, 077008.
Crouse, D.L., Pinault, L., Balram, A., Hystad, P., Peters, P.A., Chen, H., van Donkelaar, A.,
Martin, R.V., Ménard, R., Robichaud, A., 2017. Urban greenness and mortality in
Canada's largest cities: a national cohort study. The Lancet. Planet. Health 1,
e289–e297.
de Bell, S., Graham, H., Jarvis, S., White, P., 2017. The importance of nature in mediating
social and psychological benefits associated with visits to freshwater blue space.
Landsc. Urban Plan. 167, 118–127.
de Oliveira, E.S., Aspinall, P., Briggs, A., Cummins, S., Leyland, A.H., Mitchell, R., Roe, J.,
Thompson, C.W., 2013. How effective is the Forestry Commission Scotland's wood-
land improvement programme-'Woods In and Around Towns' (WIAT)- at improving
psychological well-being in deprived urban communities? A quasi-experimental
study. BMJ Open 3, 9.
de Vries, S., ten Have, M., van Dorsselaer, S., van Wezep, M., Hermans, T., de Graaf, R.,
2016. Local availability of green and blue space and prevalence of common mental
disorders in The Netherlands. Br. J. Psychiatry. Open 2, 366–372.
de Vries, S., van Dillen, S.M.E., Groenewegen, P.P., Spreeuwenberg, P., 2013. Streetscape
greenery and health: stress, social cohesion and physical activity as mediators. Soc.
Sci. Med. 94, 26–33.
Dempsey, S., Devine, M.T., Gillespie, T., Lyons, S., Nolan, A., 2018. Coastal blue space
and depression in older adults. Health Place 54, 110–117.
Department for Communities, Local Government, 2007. In: Generalised Land Use
Database Statistics for England 2005, . https://webarchive.nationalarchives.gov.uk/
20120919162210/http://www.communities.gov.uk/publications/
planningandbuilding/generalisedlanduse.
Domenech-Abella, J., Mundo, J., Leonardi, M., Chatterji, S., Tobiasz-Adamczyk, B.,
Koskinen, S., Ayuso-Mateos, J.L., Haro, J.M., 2018. The association between socio-
economic status and depression among older adults in Finland, Poland and Spain: a
comparative cross-sectional study of distinct measures and pathways. J. Affect.
Disord. 241, 311–318.
Elliott, L.R., White, M.P., Grellier, J., Rees, S.E., Waters, R.D., Fleming, L.E., 2018.
Recreational visits to marine and coastal environments in England: where, what,
who, why, and when? Mar. Policy 97, 305–314.
EuroQol, 2018. Where is EQ-5D used? https://euroqol.org/eq-5d-instruments/how-can-
eq-5d-be-used/where-is-eq-5d-used/.
EuroQol Research Foundation, 2018. Q-5D-3L user guide. pp. 34. https://euroqol.org/
publications/user-guides.
Eurostat, 2018. Methodological manual on territorial typologies: 2018 edition. In:
Angelova-Tosheva, V., Muller, O. (Eds.), . https://ec.europa.eu/eurostat/
documents/3859598/9507230/KS-GQ-18-008-EN-N.pdf/a275fd66-b56b-4ace-8666-
f39754ede66b.
Frumkin, H., Bratman, G.N., Breslow, S.J., Cochran, B., Kahn Jr., P.H., Lawler, J.J., Levin,
P.S., Tandon, P.S., Varanasi, U., Wolf, K.L., 2017. Nature contact and human health: a
research agenda. Environ. Health Perspect. 125.
Fryers, T., Brugha, T., Morgan, Z., Smith, J., Hill, T., Carta, M., Lehtinen, V., Kovess, V.,
2004. Prevalence of psychiatric disorder in Europe: the potential and reality of meta-
analysis. Soc. Psychiatry Psychiatr. Epidemiol. 39, 899–905.
Garrett, J.K., White, M.P., Huang, J., Ng, S., Hui, Z., Leung, C., Tse, L.A., Fung, F., Elliott,
L.R., Depledge, M.H., et al., 2019. Urban blue space and health and wellbeing in
Hong Kong: Results from a survey of older adults. Health Place 55, 100–110.
Gascon, M., Triguero-Mas, M., Martínez, D., Dadvand, P., Forns, J., Plasència, A.,
Nieuwenhuijsen, M.J., 2015. Mental health benefits of long-term exposure to re-
sidential green and blue spaces: a systematic review. Int. J. Environ. Res. Public
Health 12, 4354–4379.
Gascon, M., Zijlema, W., Vert, C., White, M.P., Nieuwenhuijsen, M.J., 2017. Outdoor blue
spaces, human health and well-being: a systematic review of quantitative studies. Int.
J. Hyg. Environ. Health 220 (8), 1207–1221.
Goldberg, D.P., Gater, R., Sartorius, N., Ustun, T.B., Piccinelli, M., Gureje, O., Rutter, C.,
1997. The validity of two versions of the GHQ in the WHO study of mental illness in
general health care. Psychol. Med. 27, 191–197.
Hartig, T., Mitchell, R., de Vries, S., Frumkin, H., 2014. Nature and health. In: In: Fielding,
J.E. (Ed.), Annual Review of Public Health, vol. 35. Annual Reviews, Palo Alto, pp.
207–228.
Hulme, C., Long, A.F., Kneafsey, R., Reid, G., 2004. Using the EQ-5D to assess health-
related quality of life in older people. Age Ageing 33, 504–507.
Jackson, C., 2007. The general health questionnaire. Occup. Med. 57 79-79.
Kahn, R.S., Wise, P.H., Kennedy, B.P., Kawachi, I., 2000. State income inequality,
household income, and maternal mental and physical health: cross sectional national
survey. BMJ 321, 1311–1315.
Katikireddi, S.V., Niedzwiedz, C.L., Popham, F., 2012. Trends in population mental health
before and after the 2008 recession: a repeat cross-sectional analysis of the
1991–2010 Health Surveys of England. BMJ Open 2.
Katikireddi, S.V., Niedzwiedz, C.L., Popham, F., 2016. Employment status and income as
potential mediators of educational inequalities in population mental health. Eur. J.
Public Health 26, 814–816.
Kyttä, M., Kahila, M., Broberg, A., 2011. Perceived environmental quality as an input to
urban infill policy-making. Urban Des. Int. 16, 19–35.
Lee, H.J., Dugan, E., 2015. How large is the gap between self-report and assessed mental
health and does it impact older adult mental health service utilization? J. Gerontol.
Soc. Work 58, 3–19.
Lumley, T., 2018. Package ‘survey’: analysis of complex survey samples. pp. 136. https://
cran.r-project.org/web/packages/survey/survey.pdf.
Maas, J., Verheij, R.A., de Vries, S., Spreeuwenberg, P., Schellevis, F.G., Groenewegen,
P.P., 2009. Morbidity is related to a green living environment. J. Epidemiol.
Community Health 63, 967–973.
Maas, J., Verheij, R.A., Groenewegen, P.P., de Vries, S., Spreeuwenberg, P., 2006. Green
space, urbanity, and health: how strong is the relation? J. Epidemiol. Community
Health 60, 587–592.
MacKerron, G., Mourato, S., 2013. Happiness is greater in natural environments. Glob.
Environ. Chang. 23, 992–1000.
Mann, R.E., Cheung, J.T., Ialomiteanu, A., Stoduto, G., Chan, V., Wickens, C.M., Ala-
Leppilampi, K., Goldbloom, D., Rehm, J., 2011. Estimating prevalence of anxiety and
mood disorder in survey data using the GHQ12: exploration of threshold values. Eur.
J. Psychiatry 25, 81–91.
Markevych, I., Schoierer, J., Hartig, T., Chudnovsky, A., Hystad, P., Dzhambov, A.M., de
Vries, S., Triguero-Mas, M., Brauer, M., Nieuwenhuijsen, M.J., Lupp, G., Richardson,
E.A., Astell-Burt, T., Dimitrova, D., Feng, X.Q., Sadeh, M., Standl, M., Heinrich, J.,
Fuertes, E., 2017. Exploring pathways linking greenspace to health: theoretical and
methodological guidance. Environ. Res. 158, 301–317.
McEachan, R., Prady, S., Smith, G., Fairley, L., Cabieses, B., Gidlow, C., Wright, J.,
Dadvand, P., Van Gent, D., Nieuwenhuijsen, M., 2016. The association between green
space and depressive symptoms in pregnant women: moderating roles of socio-
economic status and physical activity. J. Epidemiol. Community Health 70, 253–259.
McManus, S., Bebbington, P., Jenkins, R., Brugha, T., 2016. Mental Health and Wellbeing
in England: Adult Psychiatric Morbidity Survey 2014: a Survey Carried Out for NHS
Digital by NatCen Social Research and the. Department of Health Sciences, University
of Leicester. NHS Digital.
Mitchell, R., 2013. Is physical activity in natural environments better for mental health
than physical activity in other environments? Soc. Sci. Med. 91, 130–134.
Mitchell, R., Popham, F., 2007. Greenspace, urbanity and health: relationships in
England. J. Epidemiol. Community Health 61, 681–683.
Mitchell, R., Popham, F., 2008. Effect of exposure to natural environment on health in-
equalities: an observational population study. The Lancet 372, 1655–1660.
Mitchell, R.J., Richardson, E.A., Shortt, N.K., Pearce, J.R., 2015. Neighborhood en-
vironments and socioeconomic inequalities in mental well-being. Am. J. Prev. Med.
49, 80–84.
Morton, D., Rowland, C., Wood, C., L., M., Marston, C., Smith, G., Wadsworth, R.,
Simpson, I., 2011. Final Report for LCM2007-The New UK Land Cover Map. pp. 115.
Countryside Survey Technical Report No 11/07. http://nora.nerc.ac.uk/id/eprint/
14854/.
Noble, M., McLennan, D., Wilkinson, K., Whitworth, A., Exley, S., Barnes, H., Dibben, C.,
McLennan, D., 2007. The English Indices of Deprivation 2007.
Nutsford, D., Pearson, A.L., Kingham, S., Reitsma, F., 2016. Residential exposure to
visible blue space (but not green space) associated with lower psychological distress
in a capital city. Health Place 39, 70–78.
Park, J.J., O'Brien, L., Roe, J., Thompson, C.W., Mitchell, R., 2011. The natural outdoors
and health: assessing the value and potential contribution of secondary public data
sets in the UK to current and future knowledge. Health Place 17, 269–279.
Peen, J., Schoevers, R.A., Beekman, A.T., Dekker, J., 2010. The current status of ur-
ban‐rural differences in psychiatric disorders. Acta Psychiatr. Scand. 121, 84–93.
Pelling, M., Blackburn, S., 2014. Megacities and the Coast: Risk, Resilience and
Transformation. Routledge.
Riva, M., Bambra, C., Curtis, S., Gauvin, L., 2011. Collective resources or local social
inequalities? Examining the social determinants of mental health in rural areas. Eur.
J. Public Health 21, 197–203.
Sareen, J., Afifi, T.O., McMillan, K.A., Asmundson, G.G., 2011. Relationship between
household income and mental disorders: findings from a population-based long-
itudinal study. Arch. Gen. Psychiatr. 68, 419–427.
Semlyen, J., King, M., Varney, J., Hagger-Johnson, G., 2016. Sexual orientation and
symptoms of common mental disorder or low wellbeing: combined meta-analysis of
12 UK population health surveys. BMC Psychiatry 16, 9.
Stranges, S., Samaraweera, P.C., Taggart, F., Kandala, N.-B., Stewart-Brown, S., 2014.
Major health-related behaviours and mental well-being in the general population: the
Health Survey for England. BMJ Open 4.
Supina, A.L., Johnson, J.A., Patten, S.B., Williams, J.V.A., Maxwell, C.J., 2007. The
usefulness of the EQ-5D in differentiating among persons with major depressive
episode and anxiety. Qual. Life Res. 16, 749–754.
Taylor, M.S., Wheeler, B.W., White, M.P., Economou, T., Osborne, N.J., 2015. Research
note: urban street tree density and antidepressant prescription rates—a cross-sec-
tional study in London, UK. Landsc. Urban Plan. 136, 174–179.
Thomson, R.M., Niedzwiedz, C.L., Katikireddi, S.V., 2018. Trends in gender and socio-
economic inequalities in mental health following the Great Recession and subsequent
austerity policies: a repeat cross-sectional analysis of the Health Surveys for England.
BMJ Open 8, e022924.
Trostrup, C.H., Christiansen, A.B., Stolen, K.S., Nielsen, P.K., Stelter, R., 2019. The effect
of nature exposure on the mental health of patients: a systematic review. Qual. Life
Res. 28, 1695–1703.
Van den Berg, A.E., Maas, J., Verheij, R.A., Groenewegen, P.P., 2010. Green space as a
buffer between stressful life events and health. Soc. Sci. Med. 70, 1203–1210.
van den Berg, M., van Poppel, M., van Kamp, I., Andrusaityte, S., Balseviciene, B., Cirach,
M., Danileviciute, A., Ellis, N., Hurst, G., Masterson, D., Smith, G., Triguero-Mas, M.,
Uzdanaviciute, I., Wit, P.d., Mechelen, W.v., Gidlow, C., Grazuleviciene, R.,
Nieuwenhuijsen, M.J., Kruize, H., Maas, J., 2016. Visiting green space is associated
with mental health and vitality: a cross-sectional study in four european cities. Health
Place 38, 8–15.
Völker, S., Heiler, A., Pollmann, T., Claßen, T., Hornberg, C., Kistemann, T., 2018. Do
perceived walking distance to and use of urban blue spaces affect self-reported
physical and mental health? Urban For. Urban Green. 29, 1–9.
Volker, S., Kistemann, T., 2011. The impact of blue space on human health and well-being
- salutogenetic health effects of inland surface waters: a review. Int. J. Hyg Environ.
Health 214, 449–460.
J.K. Garrett, et al. Health and Place xxx (xxxx) xxxx
10
Ward Thompson, C., Roe, J., Aspinall, P., Mitchell, R., Clow, A., Miller, D., 2012. More
green space is linked to less stress in deprived communities: evidence from salivary
cortisol patterns. Landsc. Urban Plan. 105, 221–229.
Wheeler, B.W., Lovell, R., Higgins, S.L., White, M.P., Alcock, I., Osborne, N.J., Husk, K.,
Sabel, C.E., Depledge, M.H., 2015. Beyond greenspace: an ecological study of po-
pulation general health and indicators of natural environment type and quality. Int. J.
Health Geogr. 14, 17.
Wheeler, B.W., White, M., Stahl-Timmins, W., Depledge, M.H., 2012. Does living by the
coast improve health and wellbeing? Health Place 18 (5), 1198–1201.
White, M.P., Alcock, I., Wheeler, B.W., Depledge, M.H., 2013a. Coastal proximity, health
and well-being: results from a longitudinal panel survey. Health Place 23, 97–103.
White, M.P., Pahl, S., Ashbullby, K.J., Burton, F., Depledge, M.H., 2015. The Effects of
Exercising in Different Natural Environments on Psycho-Physiological Outcomes in
Post-Menopausal Women: A Simulation Study. Int. J. Environ. Res. Public Health 12
(9), 11929–11953.
White, M.P., Pahl, S., Ashbullby, K., Herbert, S., Depledge, M.H., 2013b. Feelings of re-
storation from recent nature visits. J. Environ. Psychol. 35, 40–51.
White, M.P., Pahl, S., Wheeler, B.W., Depledge, M.H., Fleming, L.E., 2017. Natural en-
vironments and subjective wellbeing: Different types of exposure are associated with
different aspects of wellbeing. Health Place 45, 77–84.
White, M., Smith, A., Humphryes, K., Pahl, S., Snelling, D., Depledge, M., 2010. Blue
space: The importance of water for preference, affect, and restorativeness ratings of
natural and built scenes. J. Environ. Psychol. 30 (4), 482–493.
White, M.P., Wheeler, B.W., Herbert, S., Alcock, I., Depledge, M.H., 2014. Coastal
proximity and physical activity: Is the coast an under-appreciated public health re-
source? Prev. Med. 69, 135–140.
Wood, S.L., Demougin, P.R., Higgins, S., Husk, K., Wheeler, B.W., White, M., 2016.
Exploring the relationship between childhood obesity and proximity to the coast: a
rural/urban perspective. Health Place 40, 129–136.
World Health Organisation, 2018. Mental disorders. https://www.who.int/en/news-
room/fact-sheets/detail/mental-disorders.
Wyles, K.J., Pahl, S., Thomas, K., Thompson, R.C., 2016. Factors that can undermine the
psychological benefits of coastal environments: exploring the effect of tidal state,
presence, and type of litter. Environ. Behav. 48, 1095–1126.
Wyles, K.J., White, M.P., Hattam, C., Pahl, S., King, H., Austen, M., 2017. Are some
natural environments more psychologically beneficial than others? In: The
Importance of Type and Quality on Connectedness to Nature and Psychological
Restoration. vol. 0 Environment and Behavior, 0013916517738312.
J.K. Garrett, et al. Health and Place xxx (xxxx) xxxx
11
