Maas, J., Dillen, S.M.E. van, Verheij, R.A., Groenewegen, P.P.
Social contacts as a possible
mechanism behind the relation between green space and health. Health and Place: 2009, 15(2),
586-595
Postprint 1.0
Version
Journal website [Link]
Pubmed link [Link]
DOI 10.1016/[Link].2008.09.006
This is a NIVEL certified Post Print, more info at [Link]
Social contacts as a possible mechanism behind
the relation between green space and health
JOLANDA MAASA, SONJA M.E. VAN DILLENB, ROBERT A. VERHEIJA AND PETER P.
GROENEWEGENC
a
NIVEL (Netherlands Institute for Health Services Research), P.O. Box 1568, 3500 BN
Utrecht, The Netherlands
b
CIZ, P.O. Box 232, NL-3970 AE Driebergen, The Netherlands
c
Department of Human Geography and Department of Sociology, Utrecht University, P.O.
Box 80115, NL-3508 TC Utrecht, The Netherlands
ABSTRACT
This study explored whether social contacts are an underlying mechanism
behind the relationship between green space and health. We measured social
contacts and health in 10,089 residents of the Netherlands and calculated the
percentage of green within 1 and a 3 km radius around the postal code
coordinates for each individual's address. After adjustment for socio-economic
and demographic characteristics, less green space in people's living environment
coincided with feelings of loneliness and with perceived shortage of social
support. Loneliness and perceived shortage of social support partly mediated the
relation between green space and health.
INTRODUCTION
There is increasing attention and evidence for a positive relation between the amount of
green space in the living environment and people's health and well-being. Several studies
showed that more green space in the living environment of people is positively related to
people's self-perceived health and leads to lower mortality risks (e.g. de Vries et al., 2003;
Maas et al., 2006; Mitchell and Popham, 2007; Takano et al., 2002). However, little is
known from these studies about the way in which green space exerts a beneficial effect on
health. Several mechanisms could be underlying of which the following are most commonly
mentioned: recovery from stress and attention fatigue, encouragement of physical activity
and facilitation of social contact (de Vries et al., 2003; Groenewegen et al., 2006; Health
Council of the Netherlands and Dutch Advisory Council for research on Spatial Planning,
2004; Maas et al., 2006). A large number of mainly experimental studies have produced
strong evidence of the positive effect of nature on recovery from stress and attention fatigue
(see e.g. [Hartig et al., 2003] and [Hartig et al., 1991]; Health Council of the Netherlands and
Dutch Advisory Council for research on Spatial Planning, 2004; Kaplan and Kaplan, 1989).
About other possible underlying mechanisms, such as social contacts, less is known. One
review suggests green space might ‘have beneficial effects on health in so far as green space
promotes social contact, for example through green meeting places in neighbourhoods,
group-based nature activities (e.g. walking or willow pollarding) and gardening (shared
This is a NIVEL certified Post Print, more info at [Link]
Maas, J., Dillen, S.M.E. van, Verheij, R.A., Groenewegen, P.P. Social contacts as a possible
mechanism behind the relation between green space and health. Health and Place: 2009, 15(2),
586-595
gardens for the elderly and allotment gardens)’ (Health Council of the Netherlands and
Dutch Advisory Council for research on Spatial Planning, 2004). In this study, we aim to
investigate whether social contacts are a possible mechanism behind the relation between
green space and health.
Several terms have been used in studies of the health-enhancing components of social
relationships, such as social support, social network and social integration (Berkman et al.,
2000). In this paper, we prefer to use the container term social contacts for all these terms.
Social contacts can take many forms, including having a conversation, undertaking joint
activities and paying visits. It is widely recognised that social relationships can influence a
variety of health outcomes (see e.g. Berkman et al., 2000; Hawe and Shiell, 2000). Persons
actively involved in communities or socially engaged with others tend to live longer
(Kawachi et al., 1997) and are healthier both physically as well as mentally (see e.g.
Kawachi and Berkman, 2000; Leyden, 2003).
Meeting opportunities are important for the development of local communities and social
ties with neighbours because people have to be able to meet to establish relationships (Flap
and Völker, 2005; Völker et al., 2007). Most contact between neighbours will occur in places
like local, recreation facilities, schools, churches and parks (Kuo et al., 1998; Völker et al.,
2007). The presence of trees and grass in common spaces, as opposed to barren common
spaces, may attract residents to outdoor spaces, thereby leading to more frequent contacts
among neighbours (Coley et al., 1997). Natural settings in common space are attractive
because they can for example provide shadow, privacy and sound buffering from
surrounding environments and they could have restorative effects (Coley et al., 1997; Hartig
et al., 2003; Kaplan and Kaplan, 1989). Three closely related studies, performed by the same
research group in an underprivileged area of Chicago, provide indication of a positive
relation between the presence of green public facilities and social ties (Coley et al., 1997;
Kuo et al., 1998; Kweon et al., 1998).
Coley et al. (1997) investigated the link between the use of outdoor public spaces and the
presence, number and location of trees. They found that the presence of trees ‘consistently
predicted greater use of outdoor spaces by all people, young and older, as well as groupings
of people consisting of both youth and adults together’(Coley et al., 1997). Furthermore, they
found that ‘the amount of time residents spent in common space was strongly predicted by
the presence, location and number of trees’ (Coley et al., 1997). Kuo et al. (1998) took this
line of research to the next level and studied whether greener neighbourhoods give rise to
stronger neighbourhood social ties. They found that levels of vegetation predict both use of
common spaces and the strength of neighbourhood social ties. Moreover, they found that use
of common spaces mediated the relation between vegetation and neighbourhood social ties
(Kuo et al., 1998).
Kweon et al. (1998) subsequently investigated the relationship between older adults’
exposure to nearby green common spaces and their level of social integration and attachment
to local community. They found modest relationships between the use of green outdoor
common space and the strength of neighbourhood social ties and sense of community for
older adult residents of inner-city neighbourhoods.
Overall, these studies broadened our understanding of the importance of green space for
neighbourhood social ties. However, it is unknown whether these results also apply to other
settings. The studies were conducted in highly deprived urban areas where green elements
were very scarce. The question is whether these relationships will also be found in other,
richer and greener neighbourhoods.
Besides offering meeting opportunities green spaces can also promote a general sense of
community. According to Kim and Kaplan (2004), sense of community of residents is
strengthened when they feel at home (community attachment), have bonds with others, feel a
sense of connection with the place (community identity) and have access for local
exploration (pedestrianism). An empirical investigation showed that natural features and
open spaces were the most important physical features which contribute to these four
This is a NIVEL certified Post Print, more info at [Link]
Maas, J., Dillen, S.M.E. van, Verheij, R.A., Groenewegen, P.P. Social contacts as a possible
mechanism behind the relation between green space and health. Health and Place: 2009, 15(2),
586-595
domains of sense of community. Natural features can promote a sense of community by
increasing feelings of emotional attachment to a neighbourhood and people's identity with a
place, which in turn could decrease feelings of loneliness and increases social support (Pretty
et al., 1994; Prezza et al., 2001). Therefore, it is interesting not only to investigate the direct
relation between green space and contact with neighbours but also with feelings of loneliness
and social support.
A few studies have addressed the relation between green space and social contacts and
sense of community. Flap and Völker (2005) showed that Dutch neighbourhoods with more
open green space and recreational facilities, promote a sense of community. A study by
Ewert and Heywood conducted in the US (1991) showed that undertaking activities in
natural environments appeared to have stimulating effects on social contacts and social
cohesion. The results of a study by Leyden (2003) show that people in Ireland in walkable
neighbourhoods, which are among other things characterised by the availability of local
parks, are ‘more likely to know their neighbours, to participate politically, to trust others, and
to be involved socially’ (p 1550). On the other hand, in Western Australia, Wood et al.
(2007) did not find a relation between distance to park from the respondents home and social
capital.
Overall, there are several indications of a positive relation between green space and social
contacts. The aim of this study is to investigate whether social contacts are a possible factor
mediating the relation between green space and health.
More specifically, the following questions will be answered:
1. Is the amount of green space in people's living environment related to people's health?
2. Do people with more social contacts feel healthier?
3. Is the amount of green space in people's living environment related to social contacts?
4. Can the relation between green space and social contacts explain the relation between
green space and health?
To gain more insight in the relation between social contacts and green space, we will
analyse it for different subgroups. First of all, the relation will be analysed for different age
groups to find out whether the relation is stronger for children and elderly, who have fewer
resources to develop and maintain social ties that are further away from their homes. As a
consequence of the more limited mobility of children and elderly, they rely more on nearby
neighbours and their neighbourhood to support their needs (e.g. Kweon et al., 1998).
Furthermore, the relation will be analysed for different social economic status groups to find
out whether groups with a lower social economic status benefit more from green space in
their living environment. Finally, we will investigate whether the relation is the same for
urban and rural areas. Neighbourhoods in urban areas are more likely to be deteriorated,
more often have to deal with vandalism, have a less strong local community and lower levels
of social control (Flap and Völker, 2005; Steenbekkers et al., 2006). People living in rural
areas more often socialize with their neighbours, less often experience nuisance from their
neighbours (Steenbekkers et al., 2006) and have more social contacts in general (de Jong
Gierveld, 1998). Overall, however, levels of deterioration are relatively low in the
Netherlands.
METHODS
Data
The data for this study were derived from two different datasets that were combined for this
study. The health data and data on social contacts were collected within the framework of the
second Dutch National Survey of General Practice (DNSGP-2), conducted in the
Netherlands in 2001. The DNSGP-2 included a nationwide representative sample of 104
general practices with approximately 400,000 people on their list, who were a good
representation of the Dutch population in terms of age, gender and type of health insurance
(Westert et al., 2005).
This is a NIVEL certified Post Print, more info at [Link]
Maas, J., Dillen, S.M.E. van, Verheij, R.A., Groenewegen, P.P. Social contacts as a possible
mechanism behind the relation between green space and health. Health and Place: 2009, 15(2),
586-595
As part of the DNSGP-2 a random sample of 12,699 people participated in a health
interview survey (response rate 64.5%). Questionnaires were administered by trained
interviewers during a face-to-face interview. To avoid seasonal patterns in morbidity, all
interviews were carried out within 12 months in 2001 and distributed equally across all four
seasons. People aged between 12 and 17 were interviewed, with one parent present. Privacy
of the participating persons is guaranteed and in accordance with Dutch legislation, and the
study was approved by the Dutch Data Protection Authority. Patients were informed about
the study prior to data collection (Westert et al., 2005). The socio-demographic
characteristics of the respondents were highly comparable to those of the total Dutch
population, although men, younger age groups and migrants were slightly underrepresented
(Westert et al., 2005).
Environmental data were derived from the National Land Cover Classification database
(LGN4), which contains the dominant type of land use of each 25 m×25 m grid cell in the
whole of the Netherlands in 2001 (Thunissen and De Wit, 2000). The two datasets were
matched on the basis of the x and y coordinates of the respondent's six character postal code
(on average about 15–20 households have the same six character postal code). Only
respondents who had valid responses on the relevant variables were included, leaving 10,089
respondents.
Measures
Self-reported health indicators
We used three global health indicators that have shown to be positively related with the
amount of green space in the living environment in a study by De Vries et al. (2003):
1. Perceived general health—Measured on a five-point scale, running from ‘excellent’ (1)
to ‘bad’ (5). For our purposes the scores were dichotomised with scores from good to
excellent classified as healthy (1). This kind of operationalisation has shown to be valid and
predictive of health indicators in numerous studies (Rütten et al., 2001; Simons, 2002).
2. Number of health complaints (maximum 43) experienced in the last 14 days (e.g.
headache, coughing, nausea and lower back pain) (Foets and Van der Velden, 1990).
3. Self-rated propensity to psychiatric morbidity—Measured with the Dutch 12-item
version of the General Health Questionnaire (GHQ-12) (Goldberg, 1972; Koeter and Ormel,
1991). This variable was dichotomised: scores of 2 and higher were classified as an
increased risk of psychopathology.
Social contacts
We used the following measures of social contacts:
1. Loneliness—Measured using six items that were based on the UCLA loneliness scale,
e.g. ‘I feel part of a group friends’, I feel isolated from others’, ‘There are people who really
understand me’ (Russell, 1996). A summary score counted the degree of loneliness [possible
score range 6–18], with a higher score representing a higher degree of loneliness. The scale
had an internal reliability (Cronbach's alpha) of 0.64.
2. Social support—Measured with the social support list (SSL), which has shown good
construct validity and high reliability (Bridges et al., 2002; Van Sonderen, 1993a). The SSL
measures among other things the interactions (SSL-I) and discrepancies (SSL-D) that people
experience in receiving social support from their direct environment. Each list originally
consists of 34 items which form six subscales. Only questions related to three subscales (in
total 19 items) were included in the health interview to shorten the length of the survey,
namely emotional support with problems (eight items), instrumental interactions (seven
items) and informative support (four items) (Van Sonderen, 1993b). For this study, we
investigated both the number of supportive interactions people receive (using SSL-I) and the
shortage of social support (using the SSL-D).
This is a NIVEL certified Post Print, more info at [Link]
Maas, J., Dillen, S.M.E. van, Verheij, R.A., Groenewegen, P.P. Social contacts as a possible
mechanism behind the relation between green space and health. Health and Place: 2009, 15(2),
586-595
The number of supportive interactions: The SSL-I measures the number of supportive
interactions people receive from their social support network. The response categories are as
follows: seldom/never (1); now and then (2); regularly (3); and often (4). A total score on the
SSL-I was measured by summing up the scores on the 19 items, the higher the score the
more social support (Cronbach's alpha=0.85).
Shortage of social support: The SSL-D represents the subjective experience of mismatches
between the received and desired frequency of receiving social support. The response
categories are (1) I miss it, I would like it to happen more often (2) I don’t really miss it, but
it would be nice if it happened a bit more often (3) just right, I would not want it to happen
more or less often (4) it happens too often, it would be nice if it happened less often. The
focus of this study was on the perceived lack of social support, that is ‘I miss it’. Following
the instructions for using this scale, the first category was recoded into value 3 and the
second category kept value 2. The latter two response categories (categories 3 and 4) both
received value 1. A higher score on the scale thus equals a larger shortage of social support.
A total score on the SSL-D was measured by summing up the scores on the 19 items
(Cronbach's alpha=0.86).
Only a random sample of about half the study population (n=4944) has been asked
questions concerning social support to shorten the length of the questionnaire. Furthermore,
only the respondents who filled in all questions were included (excluded 102 respondents).
3. Contact with neighbours and friends in the neighbourhood—Respondents were asked
how often they had contact with their neighbours or friends in the neighbourhood. Because
the answers were not normally distributed, we decided to make a dummy variable on
whether people often contacted neighbours and friends in the neighbourhood (more than 3
time per week or 1–3 times per week=1)) or not often (1–3 times per month, 4–11 times per
year, maximally 3 times a year=0). Only a random sample of about half the study population
(n=4635) has been asked this question in order to shorten the length of the questionnaire.
Some of the respondents (n=228) stated that this question was not applicable for them. We
left these respondents out of the analysis leaving 4407 respondents for this analysis.
Characteristics of respondents’ living environment
The following two indicators were used to measure characteristics of the respondents’
living environment:
1. Percentage of green space—Information on the environmental characteristics was
derived from the LGN4 database. The LGN4 database discriminates 39 land-use classes
including crop types, forest types, water, various urban classes and semi-natural classes and
has been proven to be valid, accurate and reliable (De Wit and Clevers, 2004; Thunissen and
De Wit, 2000). The total percentage of green space in the respondents’ living environment
was measured within 1 and 3 km radius around a respondent's home, to see whether green
space close by has a stronger or weaker effect than green space further away. The total
percentage of green space includes all urban green, agricultural green, forests and nature
conservation areas. Grid cells of 25 m×25 m are regarded as green space in the dataset if
green space has a dominant position. Gardens and small-scale green spaces, like for instance,
street trees and green roadsides are not regarded as green space in our study because they
had no dominant position in the grid cell.
2. Urbanicity—Another environmental characteristic is urbanicity. This variable consists of
five categories ranging from very strongly urban (1) to non-urban (5), and was measured at
municipal level. The indicator is based on the number of households/km2 and is widely used
in the Netherlands (Den Dulk et al., 1992).
Demographic and socio-economic characteristics
Network research reveals that social networks or social interaction of people with different
background characteristics differ considerably. Women tend to have larger and more
This is a NIVEL certified Post Print, more info at [Link]
Maas, J., Dillen, S.M.E. van, Verheij, R.A., Groenewegen, P.P. Social contacts as a possible
mechanism behind the relation between green space and health. Health and Place: 2009, 15(2),
586-595
intimate social networks (Kendler et al., 2005). The number of friends and the frequency of
social contacts decline with age (van Tilburg, 1998). Furthermore, people with higher
education and higher income levels have larger networks than people with lower education
and income levels (Tijhuis, 1994). Because social networks and health differ according to
people's background characteristics, we took several of these characteristics into account:
1. Demographic characteristics—The demographic characteristics taken into account were
gender (female=1) and age (years). To find out whether the relation between social contacts
and green space differed for different age groups, age was divided into five categories (viz.
children (aged 12–17 years), youths (aged 18–25 years), young adults (aged 26–40 years),
older adults (aged 41–65 years), elderly (aged 65+)).
2. Socio-economic status (SES) was measured by level of education (divided into four
dummy variables categories: unknown, low, middle and high) and household income
(divided into four dummy variables: income unknown, high income (net monthly income
>2450 euro), middle income (net monthly income between 1350 and 2450 euro) and low
income (net monthly income <1350 euro)). Level of education and income was categorised
because of the non-linear relation with some of the dependent variables. We included the
categories ‘unknown’ for level of education and income in order to avoid excessive drop-out
of respondents.
3. Size of the respondent's household—Because the size of the respondent's household also
influences the level of social contacts of the respondent (e.g. Tijhuis, 1994), we included a
variable measuring the size of the respondent's household in the analysis concerning the
relation between green space and the different forms of social contacts.
[TABLE 1].
Statistical analyses
The relation between the amount of green space, social contacts and health was assessed by
multilevel regression analyses, controlling for socio-economic and demographic
characteristics and urbanicity. The multilevel analyses were performed with MLwiN. We
included two levels, viz. practice and individual level, because the data of individuals were
gathered through general practices and therefore individuals clustered within practices. The
postcode level was not included because there were hardly any people from the same six
character postal code in the dataset. The independent variables, including the percentage of
green space, were centred on the average which reduces chances of multicollinearity (Draper
and Smith, 1998). The results reflect the individual level. Multilevel logistic regression
analysis was used in case the dependent variable was a dichotomy. Because we wanted to
compare the relation for different levels of urbanicity and different subgroups, we used
several models in which we included interaction effects between the urbanicity or subgroup
variable and the green indicator.
As we want to investigate whether social contacts mediate the relation between green space
and health, we used Baron and Kenny's (1986) procedure for establishing whether mediation
has occurred. To show mediation, the independent variable must significantly influence the
potential mediator, the mediator must have a significant relationship with the dependent
variable, and the relationship between the independent variable and dependent variable
should be eliminated (full mediation) or weakened (partial mediation) when the mediator is
controlled for. We used a series of multilevel (logistic) regression equations and the Sobel
test (Preacher and Hayes, 2004 K.J. Preacher and A.F. Hayes, SPSS and SAS procedures for
estimation indirect effects in simple mediation models, Behavior Research Methods,
Instruments, and Computers 34 (4) (2004), pp. 717–731. View Record in Scopus | Cited By
in Scopus (427)Preacher and Hayes, 2004) to test for mediation.
This is a NIVEL certified Post Print, more info at [Link]
Maas, J., Dillen, S.M.E. van, Verheij, R.A., Groenewegen, P.P. Social contacts as a possible
mechanism behind the relation between green space and health. Health and Place: 2009, 15(2),
586-595
RESULTS
Relation between green space and health
We investigated the relationship between green space and health both with our total sample
and with the subsample that provided data on social support and contact with neighbours.
We first investigated the relation for the percentage of green space in a 1 km radius around
people's home. Subsequently, we investigated the relation for the percentage of green space
in a 3 km radius around people's home. Overall, people with more green space in their living
environment feel healthier, have experienced a lower number of health complaints in the last
14 days and have lower self-rated propensity for psychiatric morbidity (Table 2, models 1
and 2). The relation between green space and the different health indicators was stronger and
more consistent for the percentage of green space in a 1 km radius around people's home. In
the dataset with a smaller sample size, there was no relation between the percentage of green
space in a 3 km radius around people's home and the three health indicators (Table 2, models
3 and 4).
[TABLE 2]
Relation between social contacts and health
Whether or not people often have contact with neighbours or friends in the neighbourhood
are not related to any of the health indicators (Table 2, model 5). Concerning the relation
between loneliness and health, Table 2 shows that people who feel less lonely have a better
self-perceived health, experienced fewer health complaints and have a lower self-rated
propensity for psychiatric morbidity (Table 2, model 6). People who report a larger number
of supportive interactions feel less healthy, report a higher number of health complaints and
have a higher self-rated propensity for psychiatric morbidity (Table 2, model 7). People who
experience a larger shortage of social support feel less healthy, report a higher number of
health complaints and have a higher self-rated propensity for psychiatric morbidity (Table 2,
model 8).
Relation between green space and social contacts
Loneliness
For all measures of social contact, we investigated in separate models whether they were
related to the percentage of green space in a 1 or 3 km radius. Table 3a shows that there is a
significant relation between both the percentage of green in 1 and in 3 km radii around
people's home and their feelings of loneliness. People with more green space in their living
environment feel less lonely.
[TABLE 3A AND B]
The different models in Table 4 show the results for the relation between green space and
loneliness for different subgroups. The subgroup analyses show that there is a relation
between green space and loneliness in strongly urban municipalities, for children (1 and
3 km), young adults (1 and 3 km), adults (3 km), elderly (3 km), lower educated people and
for people with a low income (1 and 3 km) (see Table 4).
[TABLE 4]
Social support
No significant relations are found between the percentage of green space in the living
environment and the number of supportive interactions respondents receive from their social
support network (see Table 3a). However, people with more green space in a 1 km radius
around their home experience less shortage of social support (see Table 3a). Our subgroup
This is a NIVEL certified Post Print, more info at [Link]
Maas, J., Dillen, S.M.E. van, Verheij, R.A., Groenewegen, P.P. Social contacts as a possible
mechanism behind the relation between green space and health. Health and Place: 2009, 15(2),
586-595
analyses show that this relation is apparent in strongly urban municipalities, for children,
youth, elderly and for people with a low income (see Table 4). There was no significant
relation between the amount of green space in a 3 km radius around people's home and
shortage of social support (see Table 3b).
Contact with neighbours or friends in the neighbourhood
Although there is no relation between whether or not people often contact neighbours and
friends in the neighbourhood and the health indicators, we did want to investigate whether
there is a relation between green space and this indicator for social contacts. There appeared
to be no significant relation between the percentage of green space and whether or not people
often contacted neighbours or friends in the neighbourhood (see Table 3).
Loneliness and shortage of social support as a possible mechanism behind the relation
between green space and health
Only for loneliness (1 and 3 km) and shortage of social support (1 km) all requirements for
testing mediation are met, since for the other mediators no relation was found with the
percentage of green space in the living environment. Therefore, we only tested whether
loneliness or shortage of social support mediate the relation between green space and the
different health indicators.
Green space, loneliness and health
Table 5 shows the results for loneliness. The results show that adding loneliness to the
model slightly influences the relation between the percentage of green space and perceived
general health, the number of health complaints in the last 14 days and people's self-rated
propensity to psychiatric morbidity. This indicates partial mediation. To test whether
loneliness mediates the relation between green space and the different health indicators we
conducted a Sobel test. Loneliness appears to partially mediate the relation between green
space and self-perceived health (1 km: z=6.26, p<0.001/3 km: z=7.43, p<0.001), number of
health complaints (1 km: z=−6.22, p<0.001/3 km: z=7.36, p<0.001) and self-reported
propensity to psychiatric morbidity (1 km: z=−5.57, p<0.001/3 km: z=6.89, p<0.001).
[TABLE 5]
Green space, shortage of social support and health
Table 6 shows the results for the triad between shortage of social support, the amount of
green space in a 1 km radius and the number of health complaints and the self-rated
propensity for psychiatric morbidity. The triad for green space, shortage of social support
and self-perceived health was not investigated, because the analyses on the relation between
green space and self-perceived health (see Table 2) showed that there was no significant
relation between the percentage of green space and self-perceived general health for the
smaller sample.
[TABLE 6]
Concerning the number of health complaints our analyses show that adding shortage of
social support to the model slightly diminishes the relation between the amount of green
space in a 1 km radius and the number of reported health complaints in the last 14 days.
Shortage of social support appears to partially mediate between the amount of green space in
a 1 km radius and number of health complaints (1 km: Sobel=−0.0039, z=−6.35,
p<0.001/3 km Sobel=−0.0036, z=5.25, p<0.001).
The analyses concerning people's self-rated propensity for psychiatric morbidity show that
adding shortage of social support to the model renders the percentage of green space in a
1 km radius insignificant. Shortage of social support mediated the relation between the
This is a NIVEL certified Post Print, more info at [Link]
Maas, J., Dillen, S.M.E. van, Verheij, R.A., Groenewegen, P.P. Social contacts as a possible
mechanism behind the relation between green space and health. Health and Place: 2009, 15(2),
586-595
percentage of green space in a 1 km radius and self-rated propensity for psychiatric
morbidity (1 km: Sobel=−0.0004, z=−6.19, p<0.001/3 km Sobel=−0.0004, z=5.30, p<0.001).
DISCUSSION
This study examined whether social contacts mediate the relation between green space and
health. We started by investigating the relation between green space and different health
indicators. Our study shows that people with more green space in 1 km radius around their
home have better self-perceived health, have experienced fewer health complaints in the last
14 days and have a lower self-rated propensity for psychiatric morbidity. This is consistent
with the studies by De Vries et al. (2003) and Maas et al. (2006), who used other datasets to
investigate the relation between green space and health, and strengthens the evidence for a
positive relation between the amount of green space in a 1 km radius around people's home
and health. However, although significant relations between the amount of green space in a
3 km radius around people's home and health have been found in the studies by De Vries et
al. (2003) and Maas et al. (2006), this study did not find a consistent significant relation,
indicating that green space closer to home might be more important for people's health.
Concerning the relation between social contacts and health our results showed that both
loneliness and experienced shortage of social support are negatively related to people's self-
perceived health, the number of health complaints and the self-rated propensity for
psychiatric morbidity. This is consistent with other studies which also find a positive relation
between social isolation, social support and health (e.g. House et al., 1988; Kawachi et al.,
1997; Kendler et al., 2005; van Oostrom et al., 1995). There was no relation between
whether or not people had frequent contacts with neighbours and any of the three health
indicators. We found a negative relation between the number of supportive interactions and
the health indicators; people with more health problems received more social support. This
negative relationship between the number of supportive interaction and health has also been
found in other studies (Tijhuis, 1994; Van Sonderen and Ormel, 1997) and may be due to the
higher need for social support of unhealthy people.
Green space and social contacts
Concerning the relation between green space and social contact, our results show that
people with more green space in their living environment feel less lonely and experience less
shortage of social support, but they did not have more contact with neighbours or friends in
the neighbourhood and they did not receive more social support. This suggests that the
relation between green space and social contacts has more to do with the fact that green
spaces can strengthen sense of community via place attachment and place identity of its
residents, than with actual contacts with neighbours. Further research is needed to investigate
whether the relation between green space and social contacts is mediated by the four
domains of sense of community.
Only green space close to home appeared to be related to social support. The subgroup
analyses show that the relation between green space and loneliness and the relation between
green space and shortage of social support was strongest in the strongly urban areas.
Furthermore, as we hypothesised, both relations appeared to be strongest for children and
elderly who, as a consequence of limited mobility rely more on their neighbourhood to
support their needs (Kweon et al., 1998). Finally, the relation was strongest for people with a
low income or a low level of education, indicating that they benefit most of green space in
their living environment for their social contacts. Apparently, the amount of green space in
the living environment is less important for social contacts of people with a high SES. The
studies conducted in Chicago also found a relation between social ties and the presence of
green public space for people with a low SES (Coley et al., 1997; Kuo et al., 1998; Kweon et
al., 1998). However, because these studies were conducted in an underprivileged area in
Chicago, no comparison could be made between the relation for lower and higher SES-
groups.
This is a NIVEL certified Post Print, more info at [Link]
Maas, J., Dillen, S.M.E. van, Verheij, R.A., Groenewegen, P.P. Social contacts as a possible
mechanism behind the relation between green space and health. Health and Place: 2009, 15(2),
586-595
A direct comparison between the results of this study and the studies conducted in Chicago
is not possible. First of all, because of the different settings in which the studies are
performed. Unlike the Chicago studies, which were performed in underprivileged areas, this
study was performed on a representative sample of the Netherlands and includes all kinds of
neighbourhoods and people with different backgrounds. Furthermore, the Chicago studies
focus on specific small-scale green space close to people's home, while in this study small-
scale green space like for instance gardens, street trees, playgrounds with trees and green
roadsides are not seen as green spaces, as a results of restrictions of the database we used.
Social contacts as an explanation for the relation between green space and health
Because only loneliness and shortage of social support were related to the amount of green
space in the living environment, we only investigated whether loneliness and a smaller
shortage of social support could explain the relation between green space and health.
We used three different health indicators to investigate this. Both loneliness and shortage of
social support partially mediated the relation between green space and the different health
indicators. Shortage of social support even completely mediated the relation between green
space and self-rated propensity for psychiatric morbidity.
Strengths and limitations
Only few studies have investigated the relation between social contacts and green space
(e.g. Ewert and Heywood, 1991; Kuo et al., 1998) and, as far as we know, there are no other
studies which also investigated the triad between green space, social contacts and different
health indicators.
The data on health and social contacts were derived from a different database then the data
on green space; consequently, there is no single source bias.
For our study, we used objective environmental measures. Objective environmental
measures reduce the risks of respondent bias. However, subjective environmental measures
can also provide important information. People's perception of green spaces may, in fact,
motivate their behaviour more than the actual amount of available green space. Unsafe green
spaces or green spaces with a low quality will be avoided by people. Thus, combining
objective measures and measures of individual's perception will improve our understanding
of how the green environment affects social contacts (Coley et al., 1997; Kuo et al., 1998;
Kweon et al., 1998).
Although the data used for this study have important advantages, they also have a few
shortcomings. First of all, our data on green space, although assessed on a small scale, does
not take small green spaces in the living environment into account. Only green space that has
a dominant position in the 25 m×25 m grid cell will be regarded as green space in the
dataset. Small-scale green spaces, which have been shown to influence the strength of
neighbourhood social ties and informal social contacts among neighbours in underprivileged
areas in Chicago, are not regarded as green space in our study because they had no dominant
position in the grid cell.
Secondly, because of the cross-sectional design of the study, it is not possible to make a
statement about the direction of causation. It is possible that people who like to have more
social contacts chose to live in greener environments because these environments offer
meeting opportunities.
A third limitation of the study is that we did not know if the social contacts of people took
place in green environments. Future research should focus on whether green space in
different environments is actually used as a place to meet others.
Some possibly important control variables could not be taken into account. Besides meeting
opportunities, several other conditions are important for the development of local
communities and social ties, namely ‘individual motivation to invest in others in the group,
alternatives to realize individual goals and interdependencies’ (Flap and Völker, 2005;
Völker et al., 2007), community attachment, community identity, social interaction,
pedestrianism (Kim and Kaplan, 2004). In this study, we only focussed on meeting
This is a NIVEL certified Post Print, more info at [Link]
Maas, J., Dillen, S.M.E. van, Verheij, R.A., Groenewegen, P.P. Social contacts as a possible
mechanism behind the relation between green space and health. Health and Place: 2009, 15(2),
586-595
opportunities in the form of a green environment. It remains unknown how the other
conditions influence the relation between green space and social contacts.
CONCLUSION
It is widely recognised that social relationships can influence several health outcomes.
Furthermore, there is increasing evidence for a positive relation between green space and
health and there are indications of a positive relation between the amount of green space in
the living environment and social relationships. However, it was unknown whether social
contacts are a possible factor mediating the relation between green space and health.
This study indicates that the amount of green space in the living environment is not only
related to people's health condition but is also positively related to people's feelings of
loneliness and shortage of social support, especially for children, elderly, and people with a
lower economic status. Furthermore, both feelings of loneliness and shortage of social
support partially mediated between the relation between green space and health.
ACKNOWLEDGMENTS
This study was supported by a grant from the Netherlands Organisation for Scientific
Research.
REFERENCES
Baron and Kenny, 1986 R.M. Baron and D.A. Kenny, The moderater-mediator variable
distinction in social psychological research: conceptual, strategic, and statistical
considerations, Journal of Personality and Social Psychology 51 (6) (1986), pp. 1173–
1182.
Berkman et al., 2000 L.F. Berkman, T. Glass, I. Brisette and T.E. Seeman, From social
integration to health. Durkheim in the new millennium, Social Science and Medicine 51
(2000), pp. 843–857.
Bridges et al., 2002 K.R. Bridges, R. Sanderman and F.L.P. Van Sonderen, An English
language version of the social support list: preliminary reliability, Psychological Reports 90
(2002), pp. 1055–1058.
Coley et al., 1997 R.L. Coley, F.E. Kuo and W.C. Sullivan, Where does community grow?
The social context created by nature in Urban Public Housing, Environment and Behaviour
29 (4) (1997), pp. 468–494.
de Jong Gierveld, 1998 J. de Jong Gierveld, A review of loneliness: concepts and definitions,
determinants and consequences, Reviews in Clinical Gerontology 8 (1998), pp. 73–80.
de Vries et al., 2003 S. de Vries, R.A. Verheij, P.P. Groenewegen and P. Spreeuwenberg,
Natural environments–healthy environments? An exploratory analysis of the relationship
between greenspace and health, Environment and Planning A 35 (2003), pp. 1717–1731.
De Wit and Clevers, 2004 A.J.W. De Wit and J.G.P.W. Clevers, Efficiency and accuracy of
per-field classification for operational crop mapping, International Journal of Remote
Sensing 25 (20) (2004), pp. 4091–4112. Den Dulk et al., 1992 C.J. Den Dulk, H. Van de
Stadt and J.M. Vliegen, Een nieuwe maatstaf voor stedelijkheid: de
omgevingsadressendichtheid, A new measure for degree of urbanisation: the address
density of the surrounding area, Maandstatistiek van de Bevolking 7 (1992), pp. 14–22.
Draper and Smith, 1998 N.R. Draper and H. Smith, III-conditioning in regression data,
Applied Regression Analysis (third ed.), Wiley Series in Probability and Statistics, New
York (1998), pp. 369–386.
Ewert and Heywood, 1991 A. Ewert and J. Heywood, Group development in the natural
environment. Expectations, outcomes and techniques, Environmental and Behaviour 23 (5)
(1991), pp. 592–[Link] and Völker, 2005 H. Flap and B. Völker, Gemeenschap,
informele controle en collectieve kwaden. In: B. Völker, Editor, Burgers in de buurt: samen
leven in school, wijk en vereniging, Amsterdam University Press, Amsterdam (2005), pp.
41–70.
Foets and Van der Velden, 1990 M. Foets and J. Van der Velden, Een nationale studie van
ziekten en verrichtingen in de huisartsenpraktijk. Basisrapport: meetinstrumenten en
This is a NIVEL certified Post Print, more info at [Link]
Maas, J., Dillen, S.M.E. van, Verheij, R.A., Groenewegen, P.P. Social contacts as a possible
mechanism behind the relation between green space and health. Health and Place: 2009, 15(2),
586-595
procedures, A national study of morbidity and interventions in general practice. Basic
report: measurement instruments and procedures, NIVEL, Utrecht (1990).
Goldberg, 1972 D.P. Goldberg, The Detection of Psychiatric Illness by Questionnaire, Oxford
University Press, London (1972).
Groenewegen et al., 2006 P.P. Groenewegen, A.E. van den Berg, S. de Vries and R.A.
Verheij, Vitamin G: effects of green space on health, well-being, and social safety, BMC
Public Health 6 (149) (2006).
Hartig et al., 1991 T. Hartig, M. Mang and G.W. Evans, Restorative effects of natural
environment experiences, Environment and Behavior 23 (1991), pp. 3–27.
Hartig et al., 2003 T. Hartig, G.W. Evans, L.D. Jamner, D.S. Davis and T. Garling, Tracking
restoration in natural and urban field settings, Journal of Environmental Psychology 23
(2003), pp. 109–123. Hawe and Shiell, 2000 P. Hawe and A. Shiell, Social capital and
health promotion: a review, Social Science and Medicine 51 (2000), pp. 871–[Link]
Council of the Netherlands and Dutch Advisory Council for Research on Spatial Planning,
2004 Health Council of the Netherlands and Dutch Advisory Council for Research on
Spatial Planning, Nature and Environment (RMNO), 2004. Nature and health. The
influence of nature on social, psychological and physical well-being. Health Council of the
Netherlands, RMNO, The Hague. (URL:Accessed 2008-05-27 (Archived by WebCite at
House et al., 1988 J.S. House, K.R. Landis and D. Umberson, Social relationships and
health, Science 214 (1988), pp. 540–545. Kaplan and Kaplan, 1989 R. Kaplan and S.
Kaplan, The Experience of Nature. A Psychological Perspective, Cambridge University
Press, Cambridge (1989).
Kawachi and Berkman, 2000 I. Kawachi and L. Berkman, Social cohesion, social capital, and
health. In: L. Berkman and I. Kawachi, Editors, Social Epidemiology, Oxford University
Press, New York (2000).
Kawachi et al., 1997 I. Kawachi, B.P. Kennedy, K. Lochner and D. Prothrow-Stith, Social
capital, income inequality, and mortality, American Journal of Public Health 87 (9) (1997),
pp. 1491–1498. )
Kendler et al., 2005 K.S. Kendler, J. Myers and C. Prescott, Sex differences in the
relationship between social support and risk for major depression: a longitudinal study of
opposite-sex twin pairs, American Journal of Psychiatry 162 (2) (2005), pp. 250–256. Kim
and Kaplan, 2004 J. Kim and R. Kaplan, Physical and psychological factors in sense of
community: new urbanist Kentlands and nearby Orchard Village, Environment and
Behavior 36 (3) (2004), pp. 313–340.
Koeter and Ormel, 1991 M.W.J. Koeter and J. Ormel, General Health Questionnaire.
Nederlandse Bewerking, General Health Questionnaire. Dutch Edition, Swets & Zeitlinger,
Lisse (1991).
Kuo et al., 1998 F.E. Kuo, W.C. Sullivan and A. Wiley, Fertile ground for community: inner-
city neighbourhood common spaces, American Journal of Community Psychology 26
(1998), pp. 823–851.
Kweon et al., 1998 B.S. Kweon, W.C. Sullivan and A. Wiley, Green common spaces and the
social integration of inner-city older adults, Environment and Behavior 30 (1998), pp. 823–
858.
Leyden, 2003 K.M. Leyden, Social capital and the built environment: the importance of
walkable neighbourhoods, American Journal of Public Health 93 (9) (2003), pp. 1546–1551
Maas et al., 2006 J. Maas, R.A. Verheij, P.P. Groenewegen, S. de Vries and P.
Spreeuwenberg, Green space, urbanity and health: how strong is the relation?, Journal of
Epidemiology and Community Health 60 (7) (2006), pp. 587–592.
Mitchell and Popham, 2007 R. Mitchell and F. Popham, Greenspace, urbanity and health:
relationships in England, Journal of Epidemiological and Community Health 61 (2007), pp.
681–683. Preacher and Hayes, 2004 K.J. Preacher and A.F. Hayes, SPSS and SAS
procedures for estimation indirect effects in simple mediation models, Behavior Research
Methods, Instruments, and Computers 34 (4) (2004), pp. 717–731.
Pretty et al., 1994 G.M.H. Pretty, L. Andrewes and C. Collett, Exploring adolescents'sense of
community and its relationship to loneliness, Journal of Community Psychology 22 (1994),
pp. 346–358.
This is a NIVEL certified Post Print, more info at [Link]
Maas, J., Dillen, S.M.E. van, Verheij, R.A., Groenewegen, P.P. Social contacts as a possible
mechanism behind the relation between green space and health. Health and Place: 2009, 15(2),
586-595
Prezza et al., 2001 M. Prezza, M. Amici, T. Roberti and G. Tedeschi, Sense of community
referred to the whole town: its relations with neighboring, loneliness, life satisfaction, and
area of residence, Journal of Community Psychology 29 (1) (2001), pp. 29–52.
Russell, 1996 D.W. Russell, UCLA loneliness scale (version 3): reliability, validity, and factor
structure, Journal of Personality Assessment 66 (1) (1996), pp. 20–40.
Rütten et al., 2001 A. Rütten, T. Abel, L. Kannas, T. Von Lengerke, G. Lüschen, J.A.
Rodriguez Diaz, J. Vinck and J. Van der Zee, Self reported physical activity, public health
and perceived environment: results from a comparative European study, Journal of
Epidemiology Community Health 55 (2001), pp. 139–146.
Simons, 2002 J.G. Simons, How is Your Health in General. Qualitative and Quantitative
Studies on Self-Assessed Health and Socioeconomic Differences Herein, Erasmus
University Rotterdam (2002).
Steenbekkers et al., 2006 C. Steenbekkers, A. Simon and V. Veldheer, Thuis op het
platteland: de leefsituatie van platteland en stad vergeleken, Sociaal Cultureel Planbureau,
Den Haag (2006).
Takano et al., 2002 T. Takano, K. Nakamura and M. Watanabe, Urban residential
environments and senior citizens’ longevity in megacity areas. The importance of walkable
green spaces, Journal of Epidemiological Community Health 56 (2002), pp. 913–918.
Thunissen and De Wit, 2000 H.A.M. Thunissen and A.J.W. De Wit, The national land cover
database of The Netherlands, ISPRS Journal of Photogrammetry and Remote Sensing 33
(2000).
Tijhuis, 1994 M.A.R. Tijhuis, Social Networks and Health, Netherlands Institute for Health
Services Research (NIVEL), Utrecht (1994).
van Oostrom et al., 1995 M.A. van Oostrom, M.A.R. Tijhuis, J.C.J.M. de Haes, R. Tempelaar
and D. Kromhout, A measurement of social support in epidemiological research: the social
experiences checklist tested in a general population in The Netherlands, Journal of
Epidemiological and Community Health 49 (1995), pp. 518–[Link] Sonderen, 1993a
F.L.P. Van Sonderen, Het meten van sociale stuen met de Sociale Steun Lijst-Interacties
(SSL-I) en Sociale Steun Lijst-Discrepanties (SSL-D): een handleiding, Measuring social
support with the social support list: a manual, Noordelijk Centrum voor
Gezondheidsvraagstukken, Groningen (1993).
Van Sonderen, 1993b Van Sonderen, F.L.P., 1993b. Social support list Van Sonderen and
Ormel, 1997 F.L.P. Van Sonderen and J. Ormel, Het meten van aspecten van sociale
steun en hun relatie met welbevinden. Een onderzoek naar de bruikbaarheid van de SSL-I
en de SSL-D, Measuring aspects of social support and their relation with well-being.
Research of the utility of the SSL-I and the SSL-D, Gedrag en Gezondheid 25 (4) (1997),
pp. 190–200.
van Tilburg, 1998 T. van Tilburg, Losing and gaining in old age: changes in personal
networks size and social support in a four-year longitudinal study, Journal of Gerontology
Series B: Psychological Sciences and Social Sciences 53 (6) (1998), pp. S313–S323.
Völker et al., 2007 B. Völker, H.D. Flap and S. Lindenberg, When are neighbourhoods
communities? Community in Dutch neighbourhoods, European Sociological Review 23 (1)
(2007), pp. 99–114.
Westert et al., 2005 G.P. Westert, F.G. Schellevis, D.H. De Bakker, P.P. Groenewegen, J.M.
Bensing and J. van der Zee, Monitoring health inequalities through general practice: the
second Dutch national survey of general practice, European Journal of Public Health 15 (1)
(2005), pp. 59–65.
Wood et al., 2007 L. Wood, T. Shannon, M. Bulsara, T. Pikora, G. McCormack and B. Giles-
Corti, The anatomy of the safe and social suburb: an explanatory study of the built
environment, social capital and residents’ perceptions of safety, Health and Place (2007).
This is a NIVEL certified Post Print, more info at [Link]
Maas, J., Dillen, S.M.E. van, Verheij, R.A., Groenewegen, P.P. Social contacts as a possible
mechanism behind the relation between green space and health. Health and Place: 2009, 15(2),
586-595
TABLES
This is a NIVEL certified Post Print, more info at [Link]
Maas, J., Dillen, S.M.E. van, Verheij, R.A., Groenewegen, P.P. Social contacts as a possible
mechanism behind the relation between green space and health. Health and Place: 2009, 15(2),
586-595
This is a NIVEL certified Post Print, more info at [Link]
Maas, J., Dillen, S.M.E. van, Verheij, R.A., Groenewegen, P.P. Social contacts as a possible
mechanism behind the relation between green space and health. Health and Place: 2009, 15(2),
586-595
This is a NIVEL certified Post Print, more info at [Link]
Maas, J., Dillen, S.M.E. van, Verheij, R.A., Groenewegen, P.P. Social contacts as a possible
mechanism behind the relation between green space and health. Health and Place: 2009, 15(2),
586-595
This is a NIVEL certified Post Print, more info at [Link]
Maas, J., Dillen, S.M.E. van, Verheij, R.A., Groenewegen, P.P. Social contacts as a possible
mechanism behind the relation between green space and health. Health and Place: 2009, 15(2),
586-595
This is a NIVEL certified Post Print, more info at [Link]
Maas, J., Dillen, S.M.E. van, Verheij, R.A., Groenewegen, P.P. Social contacts as a possible
mechanism behind the relation between green space and health. Health and Place: 2009, 15(2),
586-595
This is a NIVEL certified Post Print, more info at [Link]
Maas, J., Dillen, S.M.E. van, Verheij, R.A., Groenewegen, P.P. Social contacts as a possible
mechanism behind the relation between green space and health. Health and Place: 2009, 15(2),
586-595
This is a NIVEL certified Post Print, more info at [Link]