Document 10494

Jo urnal of Public
H~alth
I
VoL 33, No.2, pp. 212 _ 222
I
dui;lO.1093/pubmed/fcig0611
I
Ad"ao ~.c
Access
P l.lbJic~ tion
10
Septcmbe~
2010
The hea lth benefits of urban green spaces : a review
of th e eviden ce
A.C.K. lee , R. Maheswaran
S~ction of Public Healr.h, School of Health md Rciatcci R<:scarch, The University of Shcffidd, Sheffield 51 4DA, UK
Add ress correspondence to Andr<:w Le<:, E-mail: andrcud<:[email protected]<:f.ac. uk
ABSTRACT
Backg round Urban development proje<:ts. can be costly and have health impacts. An evidence-based approach t o urban planning is therefole
essential How ever, the evidence for physical and non-physical health benefits of urban green space is unclear.
Methods A literature :;earch of academic and grey literature was conducted for studies and
revl€V15.
of the health effects of green :.pace.
Articles found were appra ised for their relevance. critically leview ed and graded accordlng!y Their findings were then thematically categorized.
Resu lts There is w eak evidence for the links betw een physical, men tal health and w el!-bemg, and urban green spac.e. Environmental factors
,,
,,•
c
2
,,
such as t he quality and accessibility oj green space affect'> it s use for physical activity. User determinants. such as age, gender, ethnicity and
t he perception of safety. are also important. However, many studies w el e limited by poor study design, failwe to exclude (Onfounding, bias
or reve rs~ causality and w eak :.1atist!cal associations.
Cone/usion M ost studies reported f indings thi3t generally supported the viell\' that green space have a beneficial health effect. Establ isblng
a causal relationship is difficult, as the It'lationship is complex_ Simplistic urban interventions may therefore fail to address t he underlying
determinants of urban health that are not remediable by landscape redesign.
Keywords environment, geography, public health
Introdu ction
Globally, a dramatic demographic shift towards urbanization
is occurring. 1 Between 2000 and 2050, the proportion of
people living in urban areas is projected to rise from 46.6 to
69.6%.2 Urbanization poses problems through effects such
as environmental pollution, accidents, heat island effects and
climate change:',4 This has flagged up the need for multisectoral action to promote health in urban populations and
led to the rise of the 'Healthy Cities' movement.5,(,
Physical and psychological benefits have been linked to
green spaces through their purported effects on physical
activity.? Numerous health benefi ts of physical activity have
been documented, such as the effects on carclio- and
cerebro-vascular disease, diabetes, colorectal cancer, osteoR
porosis, depression and faU-related in juries. - 15 It also
improves mental functioning, mental health and wellbeing 16 - 22 and may have long-lasting psychological
.
benefits .2-'.' Benefits on longeVity
have also been reported.-" 4
212
i)
\Vhilst urbanization clearly has health impacts, there is
uncertainty as to whether the purported health benefits of
green spaces, such as parks and playing fields, are an urban
myth or fact. Urban developments are costly projects. It is
therefore important that urban design and planning
decisions are informed by robust evidence. This review
sought to broadly examine the evidence for the population
health benefi ts of green spaces, and to provide a narrative
summary for health policy-makers and urban planners.
M ethods
Literature searches o f electronic journal databases were conducted for stuclies and reviews of the health effects of g~een
A.C.K.Lc:e , Curucal ULtun-r 10 Public Ht!.d th
R.M:.hesw:uan , C hnk.ol Sc-ru O! Lcctuux aod I·leAd of Jh" Pubh<: H.,a]th G IS Unit
The Author 2010, Published by Oxford Uo i',«sit)' Press on beh:a.l f of F:ilCUJty of Pub lie Heal th. All righTS resrrved.
.0
"c
"
THE HEALTH BENEFITS OF URBAN GREEN SPACE
spaces. The kcyu!ords used \vere 'green space', 'public open
space', 'open space' and 'park'. The inclusion criteria were
studies and review 3.fticles referring to green or public open
spaces \.vith a health perspecti\'c, limited to human studies
and published in English . Studies and articles were excluded
if they did not pertain to health and green or public open
spaces, were published before 1990 or were purdy a descriptive or opinion piece.
In this review, the terms 'green space' and 'public open
space' were used interchangeably and presumed to be
synonymous. \~'e also looked at health effect in its broadest
sense to cover not just physical health but ,Jso mental
health and well-being. This \vas to reflect the various postulated ways in which green spaces are believed to affect
health impact sllch as through attracting people, providing
scope for physical activity to occur or having a restorative
effect. 24 - 2G In addition, we focused on articles pertaining
to high-income countries, as different contt..-xtual factors
are likely to inAuence associations seen in low- and
middle-income countries.
Databases searched included lvfedline, CINAHL, AMED,
BNI, PsycInfo, HMIr:, Cochrane library, NHS Economic
Evaluation Dat1base and the National Institute for Health
and Clinical Excellence (UK). Further back-referencing for
rdevant articles as well as an internet search for grey literature using identical terminology was also performed.
Publication searches \vere also carried out on agency \vebsites such as the Commission for Architecture and the Built
Environment (CAB E) and OPENspace, as well as UK government websites such as the Department of Health, and
Department for Culhlre, i.'vfedia and Sport (DeMS).
This literahlre review was completed in June 2010. Four
hundred and eight-five :uticles found were initially screened
for relevance. Thirty-five relevant articles were identified
and appraised for the strength and weaknesses of their
methodology and interpretations. These articles were then
graded according to the strength of evidence presented
(Tables 1- 3). Key findings from the various articles ,-vere
Table 1 Evidence grading
Evidence grode
High
InterpretatIOn of evidt;nce
The descnbed effect is plausible, preCIsely
quantified and oot vulnerable to bias
Intermediate
The descnbed effect is p!auslble but IS not
quantified precisely or may be vulnerable to bias
III
LO'N
Concerns about plausibility or vulnerability to bIas
se~ r€ty limiting thiO> value of the effect being
desCribed and quantifIed
213
then rilematic;illy summari zed and are presented in the
following section.
Results
Benefits of green space
Physical health
One postulated mechanism by \vhich green space influences
physical health is through its effect on physical activity
levels. Modification of the built environment to provide
green space offers opporhlOities for beneficial 'green exercise' such as walking."?";; Several reviews support this view
and rilere is some consensus that 'the built environment can
facilitate or constrain physical activity,.7,2G- 28 There may also
be other physical benefits, although the mechanisms for this
are not always clear. For example, rile availability of green
space has been reported to be independendy associated with
increased survival in elderly populations.z-·l- Another study
also reported a positive association between lower stroke
mortality and higher levels of greenness in the environ27
ment.
Whilst dlere is strong evidence of the health
benefits of physical activity, the evidence for the link
between physical activity levels and green space availability
1s weaker.
Mental health and wellbeing
Physical and social features of the environment may also
affect behaviour.2 ,'; Studies in various groups such as students, inner city girls and workers reported associations
between green space with a variety of psychologlcal,
emotional and mental health. bene£1ts?8,2') The provision
and access to green space also positively affects reported
stress and quality of life."o-33 A large epidemiological study
in the Nedlerlands found a positive corrdation between the
quantity of urban green space and the perception of general
health.';4 Green spaces may also influence social capital by
providing a meeting place for users to develop and maintain
35 J 7
neighbourhood social ties: The social interaction
enhances the personal and social communication skills of
"'" 1")le presence 0 f green vegetatlon
.
c
users.'
an d tlle 10rmarion of neighbourhood social ties in urban areas in turn
significantly contributes to residents' sense of safety and
3G
adjustment:
However, much of the literature on the
psychological bene6ts of green space tended to be qualitative or from grey literature sources, the quality of which
varied. Th ere is generally a lack of robust evidence for the
link between mental health, well-being and green space but
this may be due to the inherent difficulties in ·quantifying
non-physical health benefits.
~
~
~
Table 2 Studies on the relationship between green space and health
o
c:
~
Srudy
Setting
Study design
Findings
z
EVidence
>-
grade
o
~
~
~
Ba[1 E't al <11
( enn et al./lfj
Cohen et af.~o
Cohen et af.':> 2
Coombes
et al. 'j ~
Foster et aI,G1
Hillsdon
etal S!l
Hu et a1 2 ?
Kweon
et al'~~
lee et ai,'V)
Maas et a1 3 /1
45 urban neighbourhoods,
Austraha
32 urban (OmmI..lM\es,
AuS1taiid
Cross-sectional questIOnnaire survey of 1282 women. Stratified
random samphng
Cross-sectional survey of 2650 adults. Stratrt,ed cluster
sampling design
Different personal. SOCial and environmenta l factors associated
With walking for leisure.
Accessibility associated v,iith InCleased physical actiVity. Young
adults C18-3S years) reported more physical activity in the
presence of pubhc open space.
7 chies, USA
Adolescent girls liVing near parks (Within 0.5 miles) are more likely
Cross-sectional study of ' 556 adolescent girls looking at
to engage In more non-school moderate-Vigorous phYSical activity
physical activItY level~ and park u~e .
Urban setting, USA
Observational study of the usage of eight parks, Direct
More males than females use parks, and milles were WvKe as
observation at 2000 park users ilS well ilS interviEWS With
likely to be vigorously active. Residential prOXimity strongly
, 3 1B persons,
associated with park use anI;! phySICal actr",ty. People hVlng Within
a mile of a park were four times more likely to use It once a week
or more, and had 38% more ~xercise sessIons per week than
those liVing fu rthe r a~vay.
Urban setting, UK
Survey data irom 6821 adults were combined ......ith GIS ilnd
Frequency of green space use dechned with increasing distance
green space data, and analyzed.
trom the green space, Respondents livmg closer to the green
space reported higher ph~'5 lcal aCllvlty levels and W!!te les~ likely
to be obese,
ObselVational study analyzIng sUlvey lesuits for 13 927
Urban setting. UK
No correlation was found ber.veen acces.s to green spaces and
partiCIpants and GIS data.
phYSical actiVity levels,
Urban setting. UK
C ross~sectional study of 4950 respondents examlOlng access
No correlation was found between access to green spaces and
to open space and phYSICal activity,
phYSical actIVity le~ls .
Setting not stated. USA
Ecological study of sHoke mortality and da~y me tric mapPing
High levels of stroke mortality were observed In areas witillower
of air pollution and gleenness.
levels of exposure to green space.
Inner-city neighbourhood, USA Qualitative interviews orgl residential home reSidents
Exposure to green common spaces associated with better social
mtegration 01 elderly persons
82 urban neighbourhoods.
Observational ecological study comparing neighbourhood
Women With low Income or liVing In deprived neighbourhoods
USA
socioeconomic status of 2672 women and indIVidual phYSical
have less access to physical act1'Jsty resources (Intludrng parks)
actiVity.
Greater availability of phYSical ilctiVlty resources nearby appears to
benefit women Ii'ling In more deprived neighbourhoods and
IO'YV-income women more
Various settings turban. mlXcd
Self-admlllistered survey of 250782 persons of their perceived Reported that the amount of green space present m the
general health and the charactenstl{,s of their liVing
urban - rural and rural) in the
respondents' hVlng environ men~ was pOSi tIVely associated With
Netherlands
environment.
their perCEIVed general health ThiS association was 5trongN for
lower sooocconOIOl( groups, youth and the elderly.
v[ OZ '9! lplUW
uo ls:;)nS ,(q I~JO·S'r\lu.1I10rplO~XO·1l1IE:;)llQndrlf:dllll WO.lj: P:;)PIlO!Utl\On
:;;
(;
x
m
~
11
-l
J:
111
III
III
III
Maas eta/ bO
Various settings (urban. rni;o:ed
Interview$ with 4899 persons about their physl.cal actIVity.
The amount of green .space
urban-rural and rural) in the
self·perceived health, demographic and socioeconomic
related to the level of ph/slcal adrvlty undertaken by mdlViduals.
Netherlands
baCKgrounds, correlated with the quantity at green space
Iii
the living environment is sco rcely
III
avadable to each indiVidual.
Maas et aJ. N
Various settings (urban. mixed
Observational study of urban areas in Holland, comparrng
The annual prevalence rates of 15 of 24 disedse clusters wert!
urban - rural and rural) In the
proximity to green space With prevalence fates of disease using
lower In areas with more green space within a , km radiUS
Netherland:.
medical record data from 96 generai practices serving 345 143
Relationship was palticularty strong for children and tile lOwer
persons' i
socioeconomic classes. However, the effect size was small (OR:
0.95- 0 98).
Maas et aJ
)..1
Various settmgs (urban. mixed
Health Interview survey of 12,669 persons that examined
Proxrmity to green space was associated with lower ratcs of
urban - rural and rural) In the
self·reported health, social contacts, and characterrstics of the
self-reported III health, lack. of SOCial support and loneliness.
Netherlands
respondents' living environnlents.
Mitchel! and
Vari~lus
Popham JIl
urban-rural and rUfal) in
settings (urban. mixed
Observational ecological study comparing income depnvatron,
Alkause mortalrty and crrculator y dl~ease mortality was
mortality and pr oportion of green space by geographical areas
associated wIth levels of exposure to green space.
III
England
Potestlo
Urban setting. Canada
etaf W
Survey of 6772 children body-mass rndICes and their access
No dssociation was found between chIldhood obesity levels and
10 green spaces,
green space avarlability
Richardson
Small urban
Obscrvational ecological study of 1 546405 urban residents
After controlhog for confounders such as age. sex. socioeconomic
et al bJ
Zealand
in 1009 areas.
deprivation, smoking, air poHution and popUlation denSity. there
Roemmich
Setting not stated. USA
Cross·sectional analysis of a longitudinal stlldy of the
Greater
participation in physical activity of 59 children
physical activity partlcipatron by children.
Health interview survey of 11 238 respondents
Greater use of green space associated with less reported stress .
arca~ .
New
III
was no observed associatrons between green space ilnd mortality.
et af ~ I
Stigsdotter
Setting not stated. Denmark
er at-"
acc~ss
to pa rks was
ass od3t(~d
with increased levels of
III
III
J:
Closer proXimity to green space was also associated with better
'"J:
self-reported health
Sugiyama
32 urban neighbourhoods,
Cross·sect!onal mail questionnaire SUiVey of 1895 adults. Used
et af. l ~
Australia
spatlany·based sampling
Takano
et al.,101
Taylor
e('
Urban residents. Japan
at. I?
van den Berg
et af:
12
Wrtten lOt
at 82
Setting not stated, USA.
Various settings (urban, mixed
Analysed 5 year surVival of 3144 persons born
In
5 different
~
S
phYSical and mental heillth (OR' 1.37 & 1.60 respectively) as vJell
~
as recreation i:lI walkmg ,
'"
Urban areas with walkable green space assOCiated with I ncr~ased
years in 2 crties.
survival of sentor Citizens (OR' ' . 13-1 . 17).
QuestionnaIre survey of 96 parents of children With attention
Children wrth attention deficit disorder fl.JOct!on better after
defiCit disorder Convemence sampling used.
activities In green setting.
Swvey of 4529 rP.spondents
Respondents with higher levels of green space reported bel()9 less
urban-rural and rura!) rn the
affected by stressful life events, and better perceived mental
Netherlands
health.
Urban setting, !\lew Zealand
"'
Perception of neighbourhood greenne5S associated wIth better
Survey of 12 SZ9 adults correlated with GIS data on proXImity
Reported no correlatron between access to cpen spac.es and
to pa rks and beaches.
phYSical actIVIty
~
H
Z
In
."
~
v-
III
o
."
c:
~
111
~
~
Z
0~
"'"z
0,
;;:
n
n
~
'"'
I7IOZ '9 I 4::11l?W uo l s:m~ hq /~jo·sletUllOrplOJXO·4.l[Il;>4.qlldr/l:dn4. WOlJ p;>PllOjUMOO
'"
~
o
c:
~
z
>-
Table 3 Summary of reviews on green space/public open space and health
r
C
SiUdy
Settmg
Srudy design
Fmdmgs
Bauman and Bull!i"J
Predommantly North American and
RevIew of , 1 revIews of envltonmentai correlates of
Consistent asSOCiatIons bet\veen access, perceived safety and aesthetic
Australian studies.
physical actiVIty and walking.
features of parks and physical activity. limitations Identified Included lack
of standardization of measurement, wIde varrety of methods used and
rehance on cross-sectIonal study deSIgn
Bedimo-Rung,
et al.l !>
Kaczyns~l
and
Henderson"':!
Not stated .
Predominantly North American and
Australian studIes, although there were
\Jterature review of the relatronshlp between parKS,
Described health, social and economic benefits of pa rks . Proposed a
physical activity and public health to support a
conceptual model proposed.
conceptual model of the environmental attributes of a parK that aHects
RevIewed 50 quantItative studies that looked at the
For different types of parks and recreation setrings, there were dIfferent
relationship between parks and phYSical actIVity,
aSSOCiations seen, Generally, prox.imlty to parks was associated wrth
a few Eu ropean studIes CIted.
NICEn
park use.
increased phYSical actIVIty.
Vanous
UK gUIdance based on fIVe reviews examining
ModificatIon and plOmOtlon of parks may Increase walking. HO'..vever.
whether enVIronmental change affected phYSical
activity levels.
difficulties In ascribing causality to assoclatlOns. lack of eVidence, e.g. on
the long-term effect of interventions to change behaViour or of the
dlfferemiailmpact on different social group'), highlighting the need for
further research.
Morns:N
Not stated
Llteltlture review (mciudlOQ grey literature) of black
and minority ethnic groups and public open space
Identified barriers to public open space use by black and mino rity ethniC
groups
Morris7
Not stated
literature review (includlOg grey htel ature) of the
!dentlfled health, well-being, economic and social benefits of open
relationship betWeen health and open space.
space.
ReView of 18 Quantitative studIes on envrronmental
Aesthetic attnbutes and acceSSibility affected phYSical actIVity Studies
influences on walking, 16 studies used cross-sectiona l
deSign and 2 were prospective studies.
reported only a small va.riance in phYSica l actrvit~' There was also a
consistency In the patterns of associatIOns seen.
UK policy paper reviewing the determinants ot health
Reports benefits of natural settings on IndIVIdual wen-being . Also
and well-being. and connections to nature/green
exerCISe.
deSCrIbes potential publiC health benefits of Incrcaslng green exercise
Ovven et al.4;'
Pretty et al, l~
Vanous
Not stated
Transportation
Rt'scarch Board'13
Urban, USA
Travlou(,a
Vanous
Summary paper on the role of the bUilt environment
GrO'Nlng body of eVidenCe {mamly cross sectional) of aSSOCiatIon
on phYSlcal actIVity Details of methodology net
between bUilt enVlfonment and phYSlcal actIVity levels
stated .
Literature review (including grey literature) of
Described the experience and perceptions of young people With reg;;rd
teenagers and pubHc space
to public space use
T10ulas and
European polrc}' documents. Ongln of
literature revIew of both policy documents and
Varrous studies reporting a~socjauons between uroon green space and
James$ol
resea rch <\ltlcles not stated.
research articles of the role of urban green spiKe and
health and well-being Proposed that good qualJty open space is relatp.d
health .
to better Quality of lrfe ot urban residents
vi OZ; '91 q::llIlW UO ls:mg ,(q (3Jo·sIIlU.1I10rplOjXO·4 1 11:! ~ qq[ldrl/:dlll.\ WOlJ P~P\!O\UA\O(l
~
~
c
tc,
l:
iT.
~
~
THE HEALTH BENEfITS OF UR)},\N GREEN SPACE
217
SOcLoecon omic benefits of green space
Exposure to green spaces may have an impact on urban
socioeconomic health inequalities ..~8 Studies found that inner
city and poor populations are less likely to report parrici~
pation in outdoor recreation activities?6.39 Tc:enagers living
in disadvantaged neighbourhoods for e.'(ample lacked access
to parks they considered safe and were therefore less likely
to participate in phys ical activities than teens in more affluent neighbourhoods,41) Another study noted that people in
low-income households were more likely to adopt low Je\'els
of activity and \\,"Cre least well served by affordable facilities:11 Affluent residents, on the other hand, were more
likely to live in close proximity to facilities of any type,
Socioeconomic differentials in physical inactivity are consistent with socioeconomic gradients In many health outcomes and may represent a key pathway through which
socioeconomic statuS affects health. ;2 The uneCJual distribution of green space could account for some of the crosscultural and socioeconomic variations In their use, \X'hilst
access to green space appears to be implicitly linked \\,;th
levels of depr.ivation, what cannot be discounted are confounding factors such as individual lifestyles that could have
socioeconomic links.
Env ironm e nta l det erm ina nts o f phys ica l act iv ity
and g ree n space us e
The presence itself of green space is unhkely to t.'Cplain the
public health benefits suggested and the relationship is likely
to be comple..x and influenced by multiple factors including
attributes of the environment and me individuaL 3 ,26,4.3,44
Environmental influences have been identified dlat appear
to affect the use of green space and therefore lei sure ~ time
physical activity in these areas. 4 4. These include characteristics of the green space such as its features, condition,
accessibility and safery.45
Accessibili ty
NIost studies to date have consistently reported the association between ease and convenience of access with either
utilitarian forms of ph)lsical activity or leisure-time physical
.. "" ," -,. Thi S 0 bservatJon
.
actJvlty.'
app li ed botI1 to adul ts
and children,50.5 1 People with very good access to large
attractive green space were more likely to use it. Moreover,
users were also more likely to achieve recommended levels
of activity compared \\lith non-users. Residential proximity
to green spaces was also associated with increased levels of
. I acuvlty"
.. H 52 - .55. an d t,e
I presence 0 f b
·
PI'yslCa
arners
sue I1 as
major roads was an influencing factor.'w Whilst many
studies have consistently noted the importance of access
218
JOU RNAL Of PUBLIC HEALTH
and green space use, there have been exceptions. One
British study using cross-sectional methodologr failed to
demonstrate such a re1ationship:"~ Of particular note, the
authors in that study noted that positive associations
reported in other articles 'appear to be restlicted to specific
types of green spaces and walking or cycling be haviour'.
Q uality and availability of space
The quality and availability of green space may also have a
bearing on its use. 54 T his aspect includes issues of maintenance and availability of facili ties and activities that affect the
appeal of the green space.S{) People choose to use or not
use green spaces not only for its features but also the condition of those facilities and featu res. Places in disrepair are
less likely to be visited and contribute to a perceived sense
of lack of safety.26
Use r determinants of phys ical activity and gree n
s pa ce use
The personal attributes of users can affect their physical
activity levels and use of gteen space.49 They include the
following:
Age
Several studies obs erved variations in green space use by
different age groups but the findings are inconsistent. Older
persons and teenagers were commonly cited as more infrequen t users 42,57,58 but some srudies report that young adults
pattake in more leisure-time physical activi ty in the presence
of green space.4S A decline in physical activity in adolescence was also reported with total participation time in
physical activities falling by up to 37% between the ages of
15 and 18 years. 59 - 61 This trend was particularly marked for.
teenage girls.
The causes for this are not clear, although possible explanations include social exclusion, stigma, boredom, fea r of
crime or harassment, racial and ethnic tensions, heavy traffic
and litter. (,2 T he appropriateness o f the green space could
be an issue for older children who were provided \vith only
'token spaces inappropriate to their needs'. In addition, in
some areas, teenagers may experience hostile attitudes due
to an inferred association with vandalism and crime in
public space. G3 The inconsistencies in green space use by the
different age groups therefore suggest a more complex
reiMlonship.
Gender, ethnicity and disability
Gender differences in green space use were also reported.
Males used parks more than females , and were twice as
likely to be vigorously active. 52 \,{'omen were more Jikelr to
'walk purposefully rather than fo r e..xercise:' Studjes of park
use also note that eci1ruc minorities and people \vith disabil. . were 1ess like1y to use green spaces. 3fl ,1'::.;'6 - S/l ,(,4 0 ne
tOes
e.'\:planation given for these differences was the perception
of 'safety'. However, the interaction between socioeconomic
variables, gender, ethnicity and disability is complex and
confou nds associations reported. For example, women with
low income or ftom lower socioeconomic status neighbou rhoods were reported to differentially benefit from greater
physical activity resource availability.65 Furthermore, there
were few empirical studies of racial and eci11lic variation in
park lise, and much of the existing evidence was variable
and anecdotal. s(, It is therefore difficult to tease out the relative contributions o f the different fac tors implicated .
Psychological factors (e. g. self-efficacy, perceived
barriers)
Several enabling factors positively associated with increased
levels of uralking and physical activity were identified. These
include high individual motivation, positive attitude towards
the process o f being physically active and partaking in physical activity with a significanr o ciler.°6 ,67 Conversely, personal
barriers also exist such as being overweigh t, not enjoying
exercise, being too old, a lack of time due to other commitments, ill health, injury or disability or concerns about the
.
enVlronment
or unpreeli etabl e weat11er con eli'tlOns. " ''687
,
There was evidence from 14 corroborative studies that interventions were ineffective unless fundamental issues were
addressed such as individual confidence to change behaviour, cost and availability and pre ~ existing concerns of the
risks associated with walking and cycling.¥>
Safety
Several studies and surveys reported an association between
perceived safety and physical activity levels. I'l·4rJ.53.68 For
example, the stare of clistepair of green space negatively
affects its use by making it feel less safe. G9 One review
noted that safety concerns were important for children,
young people and their parents."6 The perceived safety by
women in particular was also associated \\~th levels of
walking, although there was no statistical association noted
for me n.
Limitation s of the data
A major limitation for many studies has been the preJ ominance of before-and-afrcr and cross-sectional study
design. 4s ,4H,49,7o.71 Less than 20% of studies used a co mparison group, a substantial number only measured physical
activity levels after an intervention and a minority used an
THE HEALTH BENEFITS OF URBAN GREEN SPACE
appropriate measure of physical activit;~ The follow-up
period ,-vas often short (at around 8 'weeks) and most
studies did not account for the fact that the intervention
may have only had an impact on groups tbat were already
active and not affected by the population as a whole. Many
of the studies could not e.xclude selection bias or confounding. 71) In several studies the possibility of retJ(l"Se c01JJoliry
could not be adequately excluded. For example, in studies
examining physical activity levels and proximity to green
space, it is unclear if tlus was a true association or whether
the converse applied whereby individuals \\lho were more
physically active cl10se to move into particuJar neighbour58
hoods with proximity to green space. There were also a
number of stuclies where the relationships reported were
null or not statistically signjficant. s9 - 63 There was insufficient robust evidence of a causal association bet\veen green
space and physical activity levels and it was difficult to ascertain to what extent the interventions or environmental attributes under examination were responsible for tbe cbanges
seen.72 Some of the research was based on aesthetic and
value judgements by bod1 experts and non-experts7 3 and
articles not published in peer-reviewed journals, such as
government and non-governmental documents, tended to
quote anecdotal evidence to support their conclusions.
Despite d1ese limitations, there was some consistency in
the patterns of associations reported such as the effect of
access and perception of safety on leisure-time physical
activity levels. Although many studies reported onJy a small
variance in physical activity levels, cumulatively on a
population-wide basis these could be substantial.
Furthermore, despite the limited number of gender-based
studies, strong gender differences were reported. Much of
the work has been based in American, Australian, Dutch
and British settings. In view of dle differences in ethnic
composition and socioeconomic differences between these
populations, it is unclear if findings from one urban area
can be direcdy translated elsewhere.
219
Unfortunately, physical activity levels in many developed
countries have declined over recent decades with a shift
towards more sedentary lifestyles ..J 3 Reversing this decline
could confer considerable population health benefits. 25 To
this end, the UK government set targets to increase levels
of participation in physical activity and sport including
measures for providing more cleaner, safer and greener
public spaces. 7 .1,75 The importance of creating more good
quality open space where it is lacking has also been echoed
in the Ivfarmot Review as a means of tackling health inequalities.'G However, our review has found that the evidence for
sllch policies is not strong.
Main finding of this study
Establishing a causal relationship between green spaces and
health was difficulr and reviews done so far have been
based on weak studies. Even after socioeconomic factors are
controlled for, the possibility of confounding cannot be
excluded. 77 Conducting population surveys on distinct physical healtll problems are difficult as incidence or prevalence
figures are often too low to do so and the time spans for
benefits to materialize may be long. iS Further research is
needed to quantify the strength of association between
green spaces and urban health, but also to investigate the
psycho-social and economic dimensions that are more difficult to measure. 77,79
What this study adds
That said the reported findings in stuclies were generally
consistent and supported the current view that urban design
can facilitate physical activity and reduce impediments to
exercise. Determinants such as the perception of safety, perception of attractiveness and pollution (air and noise) can
7
also be favourably changed. ;! There are also wider nonphysical benefits such as impacts on wellbeing and mental
health, as well as social inclusion. 1'>,23,25,31,."3,5S,46
limitations of this study
Discussion
What is already known on this topic
Various reviews on this topic have been carried out but
tended to be narrowly focused on a particular aspect of
health, e.g. physical or mental. OUf review sought to pull
together the evidence holistically to include all aspects of
health and well-being. Regular physical activity is important
for health and well-being and current evidence suggests
that individuals could derive bealth benefits by engaging in
as little as 30 min of moderate exercise dail)~ 70,7 1,nS
The study of ci1e determinants of urban health is complex.
Cities are constantly changing resulting in differences in
living conclitions both within and between cities. ! City-level
analysis presumes a degree of homogeneity in individual
behaviours but city-wide characteristics are not necessarily
shared by all of its inhabitants equally. The availability of
green space varies considerably ber:v,.·cen different urban
areas and no universal standards exist that detail the optimal
amount or characteristics of green space.
Assessments of the equity of access to green spaces may
be useful and tools such as geographical mapping could be
3
3
220
JOURNAL OP PUBLIC HEALTH
used for this purpose. However, spatial studies that qutl.ntify
measures such as proximity to parks poorly capture social
dimensions such as the fear of crime. There are also difficlllties capturing factors such as environmea tal barriers that
hinder access such as the presence of heavily trafficked
roads, lack of pedestrian crossings and quality of pavements.
Individual factors, such as motivation to engage in physical
activity, need addressing too. so As such. improving access
.. 1evels. S'J .81 .8"
.
. 1 aCtlVIty
a1one may not lOcrease
p h YSlca
Whilst there is some evidence and expert consensus to
suggest that g reen spaces can facilitate physical activity, the
evidence of a direct effect at present remains weak.(IO
However, the available evidence does on balance suggest a
pos itive association between green spaces and better health.
Robust research is required to firmly establish and quantify
the contribution of the different types of green spaces to
urban health, and to distinguish walking and cycling benefits
from other postulated benefits:w,s,; Prospective urban developments involving green spaces could act as 'natural experiments' and provide research opportunities to examine their
health impacts.
The relationship between green space and urban health IS
comple.x and other fac tors influence the observed associations. Health and urban planners need to be cognizant of
this complexity as simplistic interventions may fail to
address confounding factOrs, such as socioeconomic differentials whose roots are multi-faceted, that are less easily
remediable by urban landscape redesign.
Frumkin H. Urban sprnwl and public healrh.
2002;117:201- 17.
3
4
P~Ib!t;.
McMichael AJ. The urb.m mvironment :lnd health in
Health
:I.
Rep
world of
increasing gJobalisation: issues for developing counuies. BIIII World
Htalth Organ 2000;78(9):111 7 -26.
Kickbusch L Healthy cities: a working project and a b'to\\ing movement. Health Promot 1989;4(2):77 -82.
5
6
Flynn
Be.
Healthy cities: tow;ltd v.'Oddwide health promotion.
Annll Rev Pllblif Health 19%;17:299 - 309.
7
Morris N. Health, W(II-bci"f, alld opm paa: IIierabm m'i!:ll~ Edinburgh:
O PENspace, 2003.
8
Foster C, Hillsdon M, Thorogood M. Interventions for promoting
physical activity. OXMalle Do.IoboJe .~p Rep 2OO5;(1):C0003180.
001: 10.1 002/14651858.COOO3180.pub2.
9
Kahn EB, Ramsey LT, Brownson RC et ol The effectiveness of
interventions to increase physical activity- a systematic review. Am J
Pnll Med 2002;22(4S):73-107.
10 Mcisinger C. Lowe! H, H eier M rl al Association of sport:; :ilctivities
in leisure time and incident myocardial infarction in middle-age
men and women from the general population: the MONI CA/
KORA Augsburg cohort ~tudy. c ur I Cardio/'tl!f Prell &habil
2007;14(6P88- 92.
1 t Shaw KA, G ennat He, O'Rourke P et 01. Exercise for ovclwcight or
obesity. CfXhrane Dalabalt .["st &11 2006; (4) :CD003817. DOl:
I 0.1002/14651858.CD003817 .pub3.
12 Williams PT. Physical 6mess :md activity as separate he:lrt disease
risk factors: a m eta-an:llysis. M td Sri Spcl11 Exrrt 2001;33 (5):
754-62.
13 Thomas 0, Elliott EJ, Naughton GA. Exercise for type 2 di2.betes
mellitus. Codmlllt Dfllo.baIe S)'p &" 2006;(3):CD002968. 001
1O.I002/14651858.CD002968.pub2.
Acknowledgements
The authors would like to th ank Ms. Josie Messina for
her assistance with reviewing and ctitiquing the drafts of
this work.
14 Gregg EW, Pereira MA, Caspersen CJ. Physical activity, falls, and
fractures :ilffiong older adults: :t review of the epidemiologic evidence. JAm Gtr So& 2000;48(8):883- 93.
15 Gast Ge, Frenken FJ, van Leest LA et oJ. Intta-national \'ariation in
trends in overweight and leisure time physical activities in the
Netherlands since 1980: stratification according to sex, age and
urba.nisation degree. In! J O~s 2007;3 1(3):515 -20.
Funding
This review was commissioned and funded by Barking and
DagenJlam Primary Care Trust as part of an urban health
needs assessment. R.M. was the external consultam, but the
work was conducted by both amhors.
References
Galea S, Vlahov D. Urban health: evidence, challenges and directions. Annll &11 PIIVI:C Htalth 2005;26: 341-65.
2
United Natioos. lVor/d Urbanization PTOJpedl: The 2007 '&lIil;on
Puplilaholl Datavalt [onlint}. Population Division of the Deparonent
of Economic and Social Affairs, 2007. http://www.esa.un.org/
unup/ (20 November 2009, date last accessed).
16 Daley
AJ.
Exercise therapy and mental health in clinical populations:
is exercise therapy a 'worthwhile intervention? Ad" P!}l/;;aJr Tnat
2002;8,262- 70.
17 Karp A, Paillard-Borg S, WAng H (f al Mental, physical and social
components in leisure activities equally comribute to decrease
dementia risk. Demtflt Cmafr CO,gn Disorrl2006;21(2):65 - 73.
18 Sugiyama T, Leslie E, Giles-Corti B rt 0.1. Associations of neighbourhood greenness with physical and menul he:l..lth: do walking, social
coherence and local social interaction explain the re.btionships?
Epidmllol Co1l1f!1l1f11(y Htallh 2008;62:e9. doi:1O.1136/jech.2007.
064287.
J
19 Taylor AF, Kuo FE, Sullivan We. Coping with ADD- the surpris-
ing connection to green play
setting~. En~';lvlI b~ball
2001;33(1):
54- 77.
20 Gleruster D. Exercise and mental health: a review.
H ealth 1996;116(1):7 - 13.
JR
Soc Promot
THE HEALTH I:lENEl'ITS OP URBAN GREEN SPACE
21 Cr:aft LL, Landen OM. The effect of exercise on clinical depression
and depression resulting from ment.1.l illness: a mett-analysis. ] Spol1
EWIT P~Jd)oI1998;2 0 :339-57 .
38 Mitchell R, Popham E E ffect of exposure to
naC\.l~1
221
envirorunent
on health inequalities: an observational popuJation study. LAnai
2008;372:1655 - 60.
22 Lawlor DA, Hopker SW The effectiveness of exercise as an inter-
39 Hillsdon M, Lawlor DA, Ebrahim S et pI. Physical activity in older
vention in the management of depression: systematic review and
meta-regression analysis of randomized controlled trials. 8M]
position over dle life course: observations ill the British Women's
Heart and Healdt Study. ] Epidtnliol Communi!} Htolth 2008;
2001;322,1 - 8.
23 Sacker A, Cable N. Do adolescent It:.isUle-time physical activities
foster health and well-being in aduJthood? Evidence from
1~
62:344~50.
twO
40 Babey SH, Hasten TA, Brown E R. Teens living in disadvantaged
Nakamura K. \Xla('..mabe M. Urban residential enVIron·
neighbourhoods lack access to parks and get less physical activity.
Poli'.] Brief (UCLA Center for Health Polit;} Rmaf(h) 2007;
British birth cohorts. CJlTJ PlIb Ht"lth 2006;16(3):331-5.
24 Tabno
women: associations with area deprivation and with socioeconomic
ments and senior citizens' longc:vity in mc:gacity areas: the imporEpid~"'i()1
tance of walkabh: green spacc:s. ]
2002;56:913 - 8.
I'B2007(4),1 -6.
C01J1ll1unity Health
41 Panter J, Jones A, Hillsdon M. Equity of access to ph}'sical activity
facilities in an English city. Pm Mtd 2008;46:303 - 7.
of
42 BaU K, Tunperio A, Simon J tl al. Personal, social and environ-
25 Pn:tty J, Griffin M, Sellens Metal. Grnn e:xtmJe: (ompltnJelJlary Tolu
n(l/Mt, exerase ar.d dirt in p/~Ji(ai and eJ/loliollai IJ'tll-being tJlJd impilca!ionJ
for publi& heallh poli&J~ University of Essex, CES Occasional Paper
2003-1.
26 Beclimo-Rung AL, Mo\\o'"eo Aj, Cohen DA. The significance of
mental determinants of educational inequalities in walking: a multilevel study.] Epidtm;ol u",mun;!J Health 2007;61:108 - 14.
43 Transportation Research Board: Committee on
Ph~ical
Activity,
Health, Transportation and Land Use. Dotl Ihe Bllilt E nvironment
parks to physical activity and public hcalth- a conceptual modd.
Injluenct Pl!Jsiwi Amv;ty? Examining the Evidm(r--cSpeciai Rrporl 282.
Am) Prel,Mtd 2005;28 (252):159-68.
Washington, DC: Transportation Research Board, 2005.
27 Hu Z, Liebens
J,
Rao K.R. linking stroke mortality with air pol-
lution, income, and greenness in northwest Florida: an ecological
geographical study. Int] Htalth Geogr 2008;7:20.
44 Humpd N, Owcn N, Leslie E. Environmental factors associated
w:ith adults' participation in physical activity: a review. Am
J Pm
Mol 2002;22(3),188 - 99.
28 Ohm M, Miroue T, Misbima N tl aI. Effect of the physical activities
45 Owen N, Humpd N, Leslie E et 01. Understanding environmental
J Oecup
influences on ",-alking: review and research agenda. Ani J Prro MM
2004;27 (1),67-76.
Verheij RA, de Vries S et ai Morbidity is related to a green
46 National Institute fot Health and Clinical Excellence (NICE)
in leisure time and commuting to \VOrk
00
mental hea.lth.
Health 2007;49(1):46 - 52.
29
~faas],
living environment.] cpidttniol Communi!:! Health 2009;63:967 - 97.
30 National Audit O ffi ce. Enhancing Urban Grem Jp(J(e. wndon: TSO,
2006.
31 Stigsdotter UK, Ekhohn 0, Sclupperijn
J
d al Health promoting
Clinical Guidance 43. Obu;!}: The Prevention, Idmlijication, AHt!lmml
and Managemml of Olrtrweight and Obuity in Adu"! and Childrm.
London: NICE, 2006.
47 Kaczynski AT, Henderson KA. Environmental correlates of phys-
outdoor envirorunt:nts-associations berween green space, and
ical activity: a review of evidence about parks and recreation. Lei.rllre
hl:a!th, health-related quality of life and stress based on a Danish
national representative survey. J'calJd] Publif Htaltb 2010;38(4):411-7.
So 2oo7;29(4P I5-54.
32 van den Berg AB, Maas j, Verheij RA
bern'een
stressful
life
events
~t af
and
Green space as a buffer
health.
StU
Sri
Med
48 Cerin E, VandeillOotte C, Leslie E et (Il Recreational faciliti es and
J6sure-tllne phYSIcal activity: an analysis of moderators and selfefficacy
as
a
mediator.
Heallh
Pry(hol 2008;27(2,
Suppl.):
SI26- S135.
2010;70(8P203- 1O.
33 Commission for Architecruce and the Built Environment (CABE).
Durnt Parh? Dumt BthaLioSlr? Tbt Link behvun lhe Qualit! af Parh
and Uler Behaviour. wndon: CABE, 2005.
C. Moudon AV Neighbourhood design and physical activity.
Build Rn bif2008;36(5):395 - 411.
49 Lee
Verheij RA, Groenewegen PP el ai Green space, urbanity,
50 Cohen DA, Ashwood JS, Scott M et al. Public parks and physical
activity among adolescent girls. Pediotria 2006;118(5):1 381 - 9.
and health: how strong is the relation? J Epidmlio/ Community Health
51 Roemmich IN, Epstein LH, Raha S eloJ Association of access co
34 l\.faas
J,
parks and recreational facilities with the physical activity of young
2006;60587- 92.
35 Kweon BS, Sullivan WC, Wiley AR. Green common spaces and the
social
integration of inner-city
J998;30(6) ,832 - 58.
older
36 Kuo FE, Sullivan WC, Coley RL et al
adults.
Environ
Beho/)
children. Prev Med 2006;43(6):437 - 41.
52 Cohen DA, McKenzie TL, Sehgal A et al Contribution of public
parks to physical activity. Am] Pub Htalth 2007;97(3):509 - 14.
~rtile
ground for commu -
nity: inner-city neighbourhood common spaces. Am ] Community
P,],ho/ 1998;26(6) ,823 - 51.
37 Maas J, van Dillell S]\.f, Verheij RA et 01. SOCi;11 contacts as a poss-
53 Cervero R, Doncan M. WJ..lking, bicycling, and urban landscapes:
evidence from the San Francisco Bay area. Am ] PHb Health
2003;93(9),1 478 - 83.
54 Giles-Corti B, Bmomhall MH, Knuiman M d 01. Increasing
i~
ible mechanism behind the relation between green space and health.
walking-how important
Health Plae 2009;15 (2) :586-95.
public open space. Am J Pm Med 2005;28(2S2) :169 - 76.
distance to, attractiveness, and size of
a
3
g
~
222
JOURNAL OF PUB LIC HEALTH
55 Coombes E, Jones Ap, Hillsdon M. The relationship of physical
activity and overwdght to obj ectively mcasured green space accessibility and use. Sac Sci M~d 2010;70(6):816- 22.
56 Morris N. Bla(k a/ld Minon'ry EthniC" GrolljJs aJld Public OP(II Space:
Uleralurt re/JIOII. Edinburgh: OPENspace, 2003.
57 Trost SG, Owen N, Bauman AE el aI. Correlates of adults' participation in physical activity: review and update. Med Sci Sporls Exm
2002;34 J 996 - 200 1.
58 Abercrombie LC, Sallis JF, Conway TL tl al. Income and r:3cial dispanties in ::access to public parks and private recreation facilities. Am
] Pn,M,d2008;34(1),9 - 1S.
J, Foster C tI al. The relationship bet\veen
access and quality o f urban green spaa with population physical
activity. PNblit Health 2006;120(12):1127 -32.
59 Hillsdon M. Panter
60 Maas J. Verheij RA, Spteeuwenberg P et al. Ph}'Sical activity as :.I
possible medunism behind the relationship between green space
and heaJth: a multilevel analysis. BAfC Publk Heallh 2008;8:206.
61 Foster C, Hillsdon M, Jones A tt al. Objective measures of clle
environment and physical activity- results of the environment and
physical activity study in English adults. J PhyI Act Hralih
2009;6(5uppl. 1),570 - 580.
ElIl'ilTm"wtfJ Ihaf En((lIf1ufl and SN,tpfJr. PID'rirel Activi(y
NICE,200H.
London:
73 Thompson CW Ivvil:1I' of fusrar.v ill Lmlfu(ape alld lVoodland
P.:rrcptiolls,
Autmtra a/'ld
E),:p:rimct. Edinburgh: Fore~try
Commission, 1998.
74 Department of Culture. Media j.nd Sport London Strategy Unit.
Ga.me Plan: A Stralty for Drlil'l~n'I/,g GOl'ffllflltll/} Jpo/1 c"d PhyJical
Actidty Obj~rlim. London: Department for Culture, Media and
Sport, 2002.
75 D epartment of Health. CImJing Arlivi()·,' A P1!JJi,al Attil'I'ty A ction
Plan. London: D~p:Hlment of Health. 2OOS.
76 t\'Lumor M, Allen J. Goldblatt P r1 al Strol"J:k Rrvi.'1II of Htalth
[lItqHalilitJ in ElIgla/'ld Po.rt·2010 (Fair Socirty. H ralJl!J UVtJ: The Mannot
&vi~). Global Health Equity Group, University College London.
2010. http://~~ucl.ac.uk/gh eg/marmo tn:\'i ew (22 J une 10. date
last accessed).
77 Tzoulas K, Korpela K, Venn S el al. Promoting ecosystem and
human health in urban areas using Green Infrastructure: a liter:3ture
review. umdf(ape Urban Plum! 2007;81: 167 -78.
78 VCl'heij RA. Explaining urban-rural v:u;ations in healdl: a revicw of
interactions between individual and environm ent. S(}{ Sci Med
62 Potestio ~ Patel AB, Powcll CD tf aI. Is there an assllciation
between spatial access to parks/green space and childhood overweight/obesity in Calgary, Canada? Inl J B ehav Nutr Phys Ad
2009;6,77.
79 Pinder R. Kessel A. Green J It al. Exploring perceptions of health
and the environment: a qualitative study of Tbames Chase
Community Forest. HMlth Place 2009;15(1) :349-56.
63 Richardson E, Pearce J, :M:itchdl R tl al. The association bet\lo'een
green space and cause-specific mortality in urban New Zeal:.lnd: an
ecological analysis of green space utility. BMC Public Htallb
2010;10,240.
80 Cochrane T, Da\'ey RC, Gidlow C tl aI. SmaU area and individwl
level predictors of physical activity in urban communities: a multilevel study in Stoke on Trent. England. inl J E1It~'l'otl Hn PI/blic
Htallh 2009;6(2):654 - 77.
64 Aaron OJ, Stortis KL, Robertson RJ el aI. Longirudinal study of the
nwnber and choice of leisure rime physical activities from mid to late
adolescence: implications for school curricula and community cecreation programs. Arch Pedialr Adoluc Mtd 2002;156(11):1075 - 80.
1996;42(6):923- 35.
8t Kessel A, Green J, Pinder R rt cI. Multidisciplinary research in
public halth: :I. case study of research on access to green spa~.
Public Htalth 2009;123(1):32- 8.
65 Sjolie AN, Thuen F School journeys and leisure activities in rural and
urban adolescents in Norway. H talth Promot Int 2002;171:21 - 30.
82 Witten K, Hiscock R, Pearce J et aL Neighbourhood access to open
spaces and the physical activity of re~idents: a national study. PrtV
M,d 2008;47 (3),299 - 303.
66 Dovey SM, Reeder AI, Chalmers OJ. Continuity and change in
sporting and leisure time physical activity during adale~cence. Br)
Sp,", M,d 1998;32(1),53- 7.
83 Bauman AE. Bull Fe. Emironmmtai Con'datu of Pf?yJkal Ad;'!i!) and
IEValking in Adults {md Child",n: A ful!itlP oj lVl/imn. London: National
Institute for He:l.lth and Clinical Excellence, 2007.
67 Malone K Children, youth and sustainable Cities. Local Environ
2001;6(1) ,S - 12.
68 Travlou P. Ttt1lagen and Public Spact: UleraJurt nvitIU Edinburgh:
OPENspace, 2003.
69 Law M, King G, King S tf a1 P:.Irtcrns of participation in recreational and leisure activities among children with complex physical
disabilities. Dtv Mtd Child NturoI2006;48(5):337 - 42.
84 Tzoulas Fe, James P. FiJlding linkJ bel1JJt~n " rban biadiwsity ond h"man
health and Wf:1J-b~jng. 4th International Post:gr:3duat~ Research
Conference in the Built and Human Environment, 29th March 2nd April 2004. Salford. http://wwv.~ds.sa..lford.ac.uk/urbannature/
outputs/p'lpt:rs/Tzoulas_lPRC04.pdf (1 D ecember 2009, date last
accessed).
85 Department of H ealdl. At LeaJt Five a lI?«k: E videnct on the [tnpact of
Phylical Actidty mId Iff Rrltllionfhip 10 H rolth. London: Dep::artment of
Health, 2004.
70 Lee RE, Cubbin C. W1J1kleby M. Contribution of neighbourhood
socioeconomic status and physical activity resources to ph~ical
activity among women. J Epidemjol Community Heallh 2007;61:882-90.
86
7 1 Giles-Coni B. Donovan RJ. Relative influences of individual, social
environmental, and physical environmental correlates of walking.
Am J PNblic Health 2003;93(9):1583 - 9.
87 Scott D, Jackson EL. Factors that limit and stIaregies that might
72 Nadonal InstiMe for Health and Clinical Excellence (NICE) Public
Health Guidance 8. PromotiJlg and Creating Bmlt or Nullfral
Ph y~jOtl Activity Task furce. }Jt} ."'fake S((>I/alld .Arti~: A SlrolllJ for
Pb),sical Actil'iry. Edinburgh: PATI-~ 2003. http://www.scotland.gov.
uk/Publications/2003/0ZiI6324/1 7895.
encourage people's usc of public parks.
1996;14:1 - 17.
J
Pmk It:mntiOI1 A dmin
c
o
"0-•
~
3
3
~