Content uploaded by Henk Staats
Author content
All content in this area was uploaded by Henk Staats on Oct 10, 2017
Content may be subject to copyright.
Psychological Benefits of Walking: Moderation by
Company and Outdoor Environmentaphw_1051 261..280
Marcus Johansson and Terry Hartig*
Uppsala University, Sweden
Henk Staats
Leiden University, The Netherlands
Background: We aimed to assess moderation of affective and cognitive effects of
a brisk walk by urban environmental characteristics and the immediate social
context. Methods: We conducted a field experiment with time (pre-walk, post-
walk), type of environment (park, street), and social context (alone, with a
friend) as within-subjects factors. Twenty university students reported on affec-
tive states and completed a symbol-substitution test before and after each of
two 40-minute walks in each environment. The routes differed in amount of
greenery, proximity to water, and presence of traffic, buildings, and other
people. Results: On average, walking per se increased positive affect and
reduced negative affect. Feelings of time pressure declined to a greater extent
with the park walk than the street walk. Revitalisation increased during the
park walks to a greater degree when alone, but it increased more during the
walk along streets when with a friend. We found an inconclusive pattern of
results for performance on the symbol-substitution test. Conclusions: Some
psychological benefits of a brisk walk depend on the influence of the immediate
social context and features of the outdoor urban environment, including
natural features such as greenery and water.
Keywords: affect, attention, natural environment, physical activity, psycho-
logical restoration, urban environment
INTRODUCTION
Exercise invariably occurs in an environment, and environments vary in
the degree to which they support particular activities. The realisation of
* Address for correspondence: Terry Hartig, Institute for Housing and Urban Research,
Uppsala University, Box 785, SE-80129 Gävle, Sweden. Email: terry.hartig@ibf.uu.se
This study was the basis of Marcus Johansson’s master’s thesis in psychology. Marcus died
under tragic circumstances before we could complete this article as planned. This article is
dedicated to his memory and to his family. The surviving authors report no conflicts of interest.
APPLIED PSYCHOLOGY: HEALTH AND WELL-BEING, 2011, 3(3), 261–280
doi:10.1111/j.1758-0854.2011.01051.x
© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology. Published by Blackwell Publishing Ltd., 9600 Garsington
Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA.
immediate psychological benefits from exercise thus depends in part on the
environment in which it is performed. Recognising this, researchers have
sought to assess the degree of contingency. Some studies have focused on
moderation by social aspects of environments, as with comparisons of change
in affective states with exercise performed alone or with supportive others
(e.g. McAuley, Blissmer, Katula, & Duncan, 2000; Plante, Coscarelli, &
Ford, 2001). Other studies have focused on broader aspects of environments
as moderators of benefits, as with exercise indoors versus outdoors (e.g. Hug,
Hansmann, Monn, Krütli, & Seeland, 2008; Kerr et al., 2006). In the present
study, we integrate these lines of inquiry; we assess the interactive effects of
the immediate social context and the broader socio-physical environment on
changes in affective states and cognition that occur with exercise.
We focus on brisk walking as the form of exercise. It is a common type of
moderate-intensity activity that can be performed in a wide range of settings.
It imposes low demands with regard to expert ability, equipment, prepara-
tion, and knowledge. Citing a report by the US Surgeon General (USDHHS,
1996), Ekkekakis, Hall, van Landuyt, and Petruzello (2000) note that walking
is a highly popular form of health-oriented physical activity. As they dem-
onstrated in a series of experiments, brisk walks of short duration can tem-
porarily enhance subjective well-being. In that people find walking enjoyable,
they may be motivated to do it regularly, and more so than with more intense
activities. In keeping with this motivational picture, health authorities
recommend regular brisk walking as a moderate-intensity activity through
which people can maintain their physical health (e.g. Haskell et al., 2007).
For further discussion of the intensity–affect–adherence causal chain, see
Ekkekakis, Backhouse, Gray, and Lind (2008).
With its interest in the context and motivation for brisk walking, this study
addresses a significant practical concern related to urban health. The popu-
lations of many countries have increasingly concentrated in and around cities,
so that a majority of people today live in urban areas (United Nations, 2002).
Public health authorities recognise that some urban outdoor environments
discourage physical activity, including walking for recreation and transport
(Frumkin, Frank, & Jackson, 2004). Because a lack of physical activity may
manifest in poor physical and mental health, researchers have sought to
identify the characteristics of urban environments that discourage versus
promote walking. Aesthetic amenities in the residential context, such as the
presence of greenery and close proximity to parks, appear to attract people to
walking for recreation and exercise (De Vries et al., 2011). Many urbanites do
not have ready access to suitable parks, and they may have to walk for
recreation and exercise along city streets. Intervening on the immediate social
context may offset some deficits in the urban environment as a place for
walking; walking along city streets may become more attractive if one has
company (Staats & Hartig, 2004). Of interest in this regard, Wendel-Vos,
262 JOHANSSON ET AL.
© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology.
Droomers, Kremers, Brug, and van Lenthe (2007) systematically reviewed
observational studies on determinants of physical activity, and they found
that having a companion was convincingly associated with physical activities
such as neighborhood walking.
In studying the joint effects of the immediate social context and the urban
outdoor environment, we focus here on affective and cognitive changes that
are characteristic of psychological restoration. We do so for two reasons.
First, outcomes such as reduction of stress and anxiety constitute a major
category of much-studied exercise benefits with important implications for
physical and mental health in the long run (e.g. Martinsen & Raglin, 2007).
Second, complementary bodies of theory and empirical research indicate
how urban streets and parks can have opposed effects on the realisation of
restorative benefits with walking. On the one hand, research on environmental
stress indicates that common urban conditions such as noise and crowding can
overwhelm the capacity to direct attention and engender negative emotions
(e.g. Evans & Cohen, 1987). An information-overload model of the effects and
after-effects of such conditions assumes that people have a limited capacity to
focus or direct their attention, and that the demand imposed on that capacity
is a function of the complexity, predictability, and controllability of incoming
information (e.g. Cohen, 1978; Evans & Cohen, 1987; Milgram, 1970).
Because of the attentional demands of monitoring vehicles, signs, and other
people, and the aversiveness of noise, potential threats, and other conditions,
urban streets may counteract psychological benefits of walking per se.
On the other hand, research on restorative environments has addressed
how urban parks and other natural-appearing environments support restor-
ative processes. Attention restoration theory (Kaplan & Kaplan, 1989;
Kaplan, 1995) explains how the renewal of a depleted capacity to direct
attention is supported by a sense of being away from routine mental contents,
such as those related to the pursuit of goals. Attention restoration is also
thought to be promoted when exploration of the environment and aestheti-
cally pleasing features within it engage effortless attention, or fascination.
Reliance on fascination when acting obviates the need to effortfully direct
attention, allowing for rest of the inhibitory mechanism on which directed
attention depends. Kaplan and Kaplan (1989) have argued that natural
environments will in general better support being away and fascination
than other types of environments. Several studies support that claim (e.g.
Hartig, Korpela, Evans, & Gärling, 1997; Herzog, Maguire, & Nebel, 2003;
Laumann, Gärling, & Stormark, 2001). Other theory has focused on affective
responses to natural environments. Psycho-evolutionary theory (Ulrich et al.,
1991) refers to a biologically prepared tendency to respond positively to
features of nature that would have had some significance for survival during
human evolution. For a person who is experiencing stress, the sight of trees,
water, and other natural features may rapidly initiate positive affective
ENVIRONMENTAL MODERATION OF WALKING BENEFITS 263
© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology.
responding that blocks negative feelings and thoughts, thereby opening for
stress recovery. Thus, urban parks may amplify psychological benefits of
walking per se.
These theories address resource insufficiency problems that may co-occur
and indeed affect one another (e.g. Kaplan, 1995). Both describe restorative
processes sustained in part by an individual’s engagement with aesthetically
pleasing features of the environment. Guided by these theories, several field
experiments have compared the restorative effects of different environments.
The designs of these studies have varied, but they have had some important
features in common. Subjects performed a task or tasks expected to induce an
acute potential for restoration. They then spent an extended period in one of
the comparison environments. Measures of both attention and affect were
collected, just after the time spent in the environment or both before and after
the environmental “treatment”. For example, in the earliest of these studies,
subjects performed a Stroop task for 40 minutes before walking for 40
minutes in a city center or a peri-urban park or passively relaxing indoors for
a similar period (Hartig, Mang, & Evans, 1991). After the walk, those in the
park condition reported more positive affect and detected a greater percent-
age of errors in a proofreading test than those in the two other conditions.
Studies with pretest–posttest designs have also found stronger restorative
effects of walking in a park versus along streets with both attentional
and affective measures (Berman, Jonides, & Kaplan, 2008; Hartig, Evans,
Jamner, Davis, & Gärling, 2003). To our knowledge, however, no field
experiment has yet assessed moderation of this pattern of environmental
effects by the immediate social context of walking—that is, by whether one
walks alone or with company.
Results from a scenario-based experiment by Staats and Hartig (2004)
suggest how restorative benefits of walking might depend on both the imme-
diate social context and the broader environment. After viewing a series of
images suggesting a walk through a forest or a city center, their subjects
rated the attractiveness of walking for one hour in the environment shown,
either alone or in the company of a friend. The subjects also rated the
likelihood that they would realise various attentional recovery outcomes
(e.g. regaining the ability to concentrate) during such a walk. The attrac-
tiveness of each walk correlated strongly with the rated likelihood of recov-
ery. Particularly germane to the aims of the present study, the company of
a friend increased the attractiveness of the walk in the city center, but it did
not do so for the forest walk (see also Staats, van Gemerden, & Hartig,
2010). The subjects appear to have considered the forest walk better suited
for restoration when alone.
In introducing their study, Staats and Hartig (2004) proposed that the
company of a friend could enable and enhance restoration to different
degrees in different environments. A friend could enable restoration, for
264 JOHANSSON ET AL.
© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology.
example by helping to ensure safety in circumstances that, though otherwise
restorative, also involve some threat. If free access and an absence of threat
already enable restoration in an environment, then the company of a friend
could enhance restoration by making the time in the environment more
enjoyable, drawing attention away from everyday demands and stimulating
fascination, either within the social exchange or through mutual exploration
of the environment. Thus, company could enable and enhance restoration in
ways potentially distinguishable in attentional as well as affective outcomes.
As the results of Staats and Hartig suggest, however, company may not be
beneficial in all circumstances: an environment may sometimes be more
restorative for a person when he or she can enter it alone.
The present experiment follows up on the work of Staats and Hartig
(2004). Subjects took a series of four walks, on two occasions along paths in
an urban park and on two other occasions in an urban center. In each
environment they walked alone on one occasion and with a friend on the
other occasion. On each occasion they provided measures of attention and
affective states before and after the walk. We hypothesised that walking
would produce positive changes in attention and affect that are character-
istic of restoration, and that these changes would be greater when walking in
the park than along the streets. Of greatest interest was how, if at all, the
immediate social context would modify the broader environmental effects.
The results from the scenario study by Staats and Hartig suggested that
having the company of a friend would prove more beneficial in the urban
environment.
METHOD
Design
The experimental design included time of measurement (pre-walk, post-
walk), urban outdoor environment (park, P; street, S), and immediate social
context (with or without a friend) as within-subjects factors and gender as a
between-subjects factor. Each participant alternated between the two envi-
ronment conditions over the four testing occasions (P-S-P-S or S-P-S-P), with
assignment to a given order at random but balanced across subjects. We
included gender in the design because men and women may differ in their
sensitivity to the exercise environment (Hassmén, 1996).
Subjects
Ten women (age M=22.4, SD =3.03) and 10 men (age M=24.2, SD =2.81)
between 20 and 29 years of age participated for payment (400 Swedish
crowns; approx. 60 USD). All of the subjects were students at a large Swedish
ENVIRONMENTAL MODERATION OF WALKING BENEFITS 265
© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology.
university. Their study/work load varied between 2 and 8 hours per day
(M=4.7, SD =1.96). They exercised between 0 and 13 hours per week
(M=4.55, SD =3.56). Inclusion criteria included freedom from problems
with pain, asthma, or any other condition that would hinder the ability to
walk at a brisk pace for up to one hour. The conduct of the study complied
with the legal requirements of Sweden regarding research ethics. The subjects
were informed about the demands of the experiment prior to consenting to
participate.
Outdoor Environments and Walking Routes
All testing took place in a simply furnished room in the building that served
as the point of departure and return for both of the two routes. The subjects
walked the given route for 20 minutes, then turned and walked back along
the same route. Both routes went over flat ground without obstacles. Both
routes passed through safe urban areas familiar to the participants, and they
were easy to follow. All of the subjects received a map that showed the
routes.
The street walk followed sidewalks used by many pedestrians. The streets
themselves had two lanes and, at the times scheduled for the walks, typically
carried vehicular traffic of moderate intensity; vehicles would pass regularly
at a low speed, but without traffic backing up, except at traffic lights. The
route out had two street crossings with traffic lights and three without them;
the subjects thus made 10 crossings in all, out and back. For half of the walk,
the route passed low-rise stores and office buildings. During the other half it
passed mainly 2–5-storey apartment buildings. In addition to their height and
function, the buildings show diversity in historical period, architectural style,
and color. Trees and vegetation could be seen here and there, but were in
general scarce.
The park walk took place in a municipal park across the street from the
building where all testing took place. The route followed a gravel road along
a river, first passing through an allée of hardwood trees in a landscaped area
with bushes and grass, a small pond, wooden buildings and other gravel
paths. The route continued past a small athletics facility, a mini-golf course,
and a large open field. The subjects could see boats along the river and an
industrial area on its other side, and they could see and hear traffic from
nearby roads. They did not, however, approach roadways during the walk,
apart from crossing the street on entering and exiting the park.
During the period for data collection the weather conditions remained
stable, without substantial rain or snowfall and with the temperature between
-3 and +7 degrees Celsius. Of the days of data collection, roughly half had
clear to half-clear weather, and the remainder had more substantial cloudi-
ness or heavy cloud cover.
266 JOHANSSON ET AL.
© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology.
Dependent Measures
Affective States. We used the Revitalisation, Positive Engagement,
Tranquillity and Physical Exhaustion subscales from the Exercise-Induced
Feeling Inventory (EFI; Gauvin & Rejeski, 1993) to measure feelings com-
monly affected by exercise. The form was previously back-translated from
English to Swedish (Bodin & Hartig, 2003). Subjects rate the extent to which
given words (e.g. calm, relaxed, revived, refreshed) describe how they feel at
the given moment (0 =do not feel;4=feel very much). Internal consistency
(Cronbach’s a) varied across the eight test administrations as follows: for
Revitalisation, .25 to .80; for Positive Engagement, .69 to .94; for Tranquil-
lity, .71 to .92; for Physical Exhaustion, .65 to .81. We could not identify any
one Revitalisation item that could be dropped to enhance alpha across the
eight test administrations, including those four for which alpha fell below .60.
We retained the measure despite the sometimes weak internal consistencies,
given the theoretical relevance of the construct. We used the mean of the
responses to the three items in each subscale as the variable for analysis.
We used the Negative Mood Scale (NMS; Scott, Brandberg, & Öhman,
2001) to measure negative affects. Following the guidance given by the
authors of the scale (p. 3), we used instructions suited to the measurement of
immediate rather than more persistent affective responses. Subjects rate the
extent to which given words (e.g. worried, anxious, sad, angry, irritable,
rushed) describe how they feel at the moment (0 =not at all;4=very much).
The internal consistencies were uniformly high across test administrations for
the eight-item Anxiety/Depression subscale (a=.82–.93) but sometimes low
for the four-item Anger subscale (a=.43–.83). Alphas for the four-item Time
Pressure subscale varied from -.06 to .78, but after dropping one item, they
ranged from .32 to .84. The occasional low alphas for Anger and Time
Pressure appear to have followed from limited variability in the constituent
items (which are reflected in the subscale means and standard deviations
reported in the Results section). We used the mean of the responses to the
items as the variable for analysis.
Attention. The subjects completed a behavioral measure of attention just
before and immediately after each walk. A symbol substitution test (SST)
requires sustained directed attention and directed visual shifting (e.g. Lezak,
1983). A subject writes numbers into a series of blanks, each of which is
paired with one of nine symbols. The appropriate number for a symbol is
indicated by a key. After a practice trial, the subject is given 60 seconds to fill
in as many of the 110 available blanks as possible. The score is the number of
correctly assigned numbers. We constructed four versions of the test by
changing symbol-number combinations in the key. We varied the order of
completion for the different versions randomly across test occasions.
ENVIRONMENTAL MODERATION OF WALKING BENEFITS 267
© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology.
Background Measures
For people who have no need for restoration it would be inappropriate to
characterise any benefits of exercise as restorative. In addition to the pre-walk
measures obtained with the EFI and NMS, which tapped current affective
states just before walking, we used two measures to gauge the subjects’ need
for restoration at the time of their participation. The two additional measures
tapped into persistent experiences of stress and cognitive fatigue over
extended periods. The use of these measures is an alternative to the induction
of acute restoration needs with demanding tasks just before the period spent
in the given environment (e.g. Hartig et al., 1991; Laumann, Gärling, &
Stormark, 2003; cf. Bodin & Hartig, 2003).
Perceived Stress. Participants completed the Perceived Stress Scale
(PSS; Cohen & Williamson, 1988) on the first and last test occasions. The PSS
is a global measure with 10 items that reflect on overload and a lack of
predictability and control in life situations (e.g. “In the last month, how often
have you felt difficulties were piling up so high that you could not overcome
them?”). Items are scored on a 5-point scale (0 =never;4=very often). After
reversal of positively worded items, the total score can vary between 0 and 40,
with 40 indicating a high level of global perceived stress. In the present
sample, awas .81 on the first test occasion and .78 on the second test
occasion. Note that, with instructions that referred to the subject’s experience
over the last month, the second test administration covered the subjects’
general experience of stress during the 4-week period of participation in the
study.
Cognitive Failures. Subjects completed the Cognitive Failures Question-
naire (CFQ; Broadbent, Cooper, FitzGerald, & Parkes, 1982) on the first test
occasion only. The CFQ consists of 25 questions about the frequency with
which various common mistakes in perception, memory, and action have
occurred during the past 6 months. Its authors propose that the score indi-
cates a person’s degree of vulnerability to stress (or existing attentional
fatigue). Responses are made on a scale ranging from never (=0) to very often
(=4). Scores may thus range from zero to 100. In the present sample, a=.69.
Both the CFQ and the PSS were back-translated from English to Swedish.
Procedure
Data collection took place on weekdays during a 5-week period (early Feb-
ruary to mid-March). To balance variation in weather across the four experi-
mental conditions, four participants were scheduled to walk each day, one in
each condition and with staggered departures (12:00, 12:25, 14:00, 14:25).
268 JOHANSSON ET AL.
© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology.
With the given departure times, the subjects could complete all of the walks
in early afternoon daylight. They completed the four walks at one-week
intervals.
On two specified occasions the participants were instructed to bring along
a friend. The friend could be of the same or different gender; the only
requirement was that the person not be someone with whom the subject was
romantically involved. The friend was only along as company on the walk
and was not asked to provide data. All walks were to proceed at what for the
subject was a brisk pace, without stopping.
With each of the four test occasions, the subjects completed a set of
measures before and after the walk. On the first test occasion, before the
walk, the participants performed the symbol substitution task (SST), pro-
vided background information (age, gender, etc.), and then completed the
CFQ, PSS, NMS, and EFI, in that order. Immediately after the walk the
subjects again completed the SST, NMS, and EFI. On the three subsequent
test occasions the subjects completed the SST, NMS, and EFI before the
walk and then again immediately after the walk. On the final occasion, after
the fourth walk, the subjects completed the SST, NMS, and EFI as usual.
They then also completed the PSS for the second time. Finally, they ranked
their preferences for the four conditions (i.e. park walk, alone and together;
street walk, alone and together; from 1 for most preferred to 4 for least
preferred).
Statistical Analysis
We assessed the effects of our experimental manipulations with repeated-
measures ANOVA that included time (pre-walk, post-walk), outdoor
environment (park, street), and social context (alone, with a friend) as within-
subjects factors and gender as the sole between-subjects factor. The contrasts
of focal concern were for the main effect of time and the Time ¥Environment,
Time ¥Social Context, and Time ¥Environment ¥Social Context interac-
tions. We examined whether change over time in the outdoor environment
and/or social context further depended on gender; however, we observed no
statistically significant moderating effects of gender on change during the
walk, and the effect sizes were generally modest. For the sake of simplicity, we
do not comment on them here. Given the uncertainty regarding the character
of possible high-order interactions, we assessed statistical significance with
two-tailed probabilities (a=.05). We used bivariate correlations (Pearson’s r
or Spearman’s r, depending on the character of the distributions) to deter-
mine whether the potential for psychological benefits from the walks
co-varied with the need for restoration as measured with the PSS and CFQ.
We used non-parametric tests to compare the preferences rankings for the
different experimental conditions.
ENVIRONMENTAL MODERATION OF WALKING BENEFITS 269
© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology.
RESULTS
Effects of Walking, Outdoor Environment, and Social
Context on Affective States
Just before each walk the subjects reported low-to-moderate levels of the
positive affects and low levels of the negative affects (see Table 1). As
reflected in the main effects of time (see Table 2), walking per se had mod-
erate to large, statistically significant effects on the affect measures, increas-
ing the positive affects and, despite their low initial values, decreasing the
negative ones. Only for Physical Exhaustion was the main effect of time not
significant.
Of the positive affects measured, only Revitalisation showed a significant
effect of the main experimental manipulations. The degree of change in
Revitalisation over the course of the walk apparently depended upon both
the outdoor environment and the social context. When walking in the park,
Revitalisation increased to a greater degree while alone than when with a
friend. When walking along streets, however, Revitalisation increased while
walking with a friend but it apparently did not increase while walking alone.
TABLE 1
Means and Standard Deviations for Exercise-Induced Feelings and Negative
Moods as a Function of Outdoor Environment, Social Context, and
Measurement Point
Measure
Park Street
Alone Friend Alone Friend
Pre Post Pre Post Pre Post Pre Post
Revitalisation M1.25 1.72 1.55 1.85 1.42 1.40 1.17 1.62
SD 0.73 0.80 0.76 1.00 0.97 0.96 0.63 0.74
Positive Engagement M1.68 1.98 1.93 2.20 1.67 1.65 1.48 1.90
SD 0.83 0.78 1.01 1.01 1.04 1.07 0.75 0.68
Tranquillity M1.68 2.15 1.88 2.28 1.95 1.85 1.70 1.77
SD 0.87 0.67 0.95 0.82 0.91 0.88 0.86 0.88
Physical Exhaustion M0.93 1.08 0.77 0.78 0.82 1.03 1.20 0.72
SD 0.78 0.97 0.73 0.84 0.70 0.90 0.93 0.66
Anxiety/Depression M0.69 0.39 0.49 0.26 0.49 0.38 0.55 0.28
SD 0.79 0.54 0.58 0.38 0.63 0.63 0.56 0.37
Anger M0.45 0.20 0.34 0.18 0.39 0.24 0.43 0.25
SD 0.68 0.37 0.40 0.28 0.44 0.41 0.49 0.29
Time Pressure M0.82 0.40 0.77 0.42 0.57 0.53 0.63 0.48
SD 0.70 0.71 0.73 0.55 0.53 0.55 0.44 0.57
Note: Scores can range from 0 to 4, with higher values indicating higher levels.
270 JOHANSSON ET AL.
© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology.
The outdoor environment and social context manipulations also had
mixed effects on the negative affects. Change in Physical Exhaustion was
significantly modified by the social context but not by the environment;
feelings of exhaustion increased when the subjects walked alone but not when
they walked with a friend. In contrast, change in Time Pressure was signifi-
cantly modified by the environment but not by the social context; feelings of
being rushed declined significantly more during the walk in the park than
during the walk along city streets. For none of the negative affects was change
significantly modified by both environment and social context. Note,
however, that the pattern of means for Anxiety/Depression mirrors that for
TABLE 2
Experimental Effects on Affective States
Dependent measure Effect F(1, 18) pⱕPartial h2
Revitalisation T 8.34 .010 .317
T¥E 1.65 .216 .084
T¥SC 1.12 .304 .059
T¥E¥SC 6.63 .019 .269
Positive Engagement T 14.17 .001 .441
T¥E 0.28 .605 .015
T¥SC 1.66 .214 .084
T¥E¥SC 2.65 .121 .128
Tranquillity T 7.14 .016 .284
T¥E 3.73 .069 .172
T¥SC 0.10 .755 .006
T¥E¥SC 0.45 .511 .024
Physical Exhaustion T 0.04 .845 .002
T¥E 2.20 .155 .109
T¥SC 6.08 .024 .253
T¥E¥SC 2.82 .110 .135
Anxiety/Depression T 41.38 .001 .697
T¥E 1.30 .270 .067
T¥SC 0.36 .556 .020
T¥E¥SC 3.40 .082 .159
Anger T 12.79 .002 .415
T¥E 0.20 .662 .011
T¥SC 0.11 .743 .006
T¥E¥SC 0.33 .574 .018
Time PressureaT 5.66 .029 .239
T¥E 6.14 .023 .254
T¥SC 0.04 .849 .002
T¥E¥SC 2.00 .175 .100
Note:T=main effect of time; T ¥E=Time ¥Environment interaction; T ¥SC =Time ¥Social Context
interaction; T ¥E¥SC =Time ¥Environment ¥Social Context interaction.
aAnalyses used log10 transformed variables to mitigate positive skew. The means reported in Table 1 are for
the untransformed variables.
ENVIRONMENTAL MODERATION OF WALKING BENEFITS 271
© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology.
Revitalisation; in the park, Anxiety/Depression appears to have decreased
to a greater degree while walking alone than while walking with a friend,
whereas along the streets it declined to a greater degree while walking with a
friend than while walking alone. Again, though, while these results mirror
those seen with Revitalisation, the effect is smaller and not significant.
We checked our conclusions concerning the negative affects, given the
strongly skewed distributions for those scores. We mitigated the positive
skew in the Physical Exhaustion, Anxiety/Depression, Anger, and Time
Pressure scores with logarithmic transformations (Tabachnik & Fidell, 2001)
and then repeated the analyses. Compliance with assumptions of the
RM-ANOVA, conclusions regarding statistical significance, and effect size
estimates remained much the same with the transformed variables, with one
exception. For Time Pressure, the analysis with log10 transformed variables
produced notably better results for the tests of equality of covariance matri-
ces and error variances. The conclusions regarding statistical significance
were the same with the transformed and untransformed variables for Time
Pressure, but the F-statistic and accompanying effect size estimate (partial h2)
for the effect of time were roughly 50 per cent smaller with the transformed
variables. In Table 2 we report the (more conservative) test results based on
the transformed variables; however, we report the means and standard devia-
tions for the untransformed Time Pressure variables in Table 1.
Effects of Walking, Outdoor Environment, and Social
Context on Attention
The subjects made roughly 36–40 correct number-to-symbol matches on
average with each of the administrations of the symbol substitution test (see
Table 3). Walking per se appears to have caused a slight though significant
TABLE 3
Means and Standard Deviations for Performance on the Symbol-Substitution
Test as a Function of Outdoor Environment, Social Context, and Measurement
Point
Measure
Park Street
Alone Friend Alone Friend
Pre Post Pre Post Pre Post Pre Post
Symbol-substitution M38.65 37.85 40.00 36.35 37.85 37.80 37.70 36.85
SD 5.28 5.10 6.78 5.09 5.21 4.87 4.78 4.79
Note: Scores represent the number of correct number-to-symbol matches made within 60 seconds.
272 JOHANSSON ET AL.
© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology.
decline in performance; for the main effect of time, F(1, 18) =4.75, p=.043,
partial h2=.209. The decline was greater with the walks in the park than
with the two street walks; for the Time ¥Environment interaction, F=5.99,
p=.025, partial h2=.250. Both of these effects were unexpected. Note,
however, that the effects did not involve a difference in SST performance
opening up over time. Rather, they involved an initial difference closing;
performance was on average significantly better before the park walks,
t(19) =2.53, p=.02. We have no explanation for this initial difference.
The scores are approximately normally distributed for each of the test
administrations (for all eight Kolmogorov-Smirnov tests, p>.17), and
no value exceeds 3 ⫾SDs from the respective mean. The effect appears
to reflect regression to the mean. The interaction of time and social con-
text was not significant (F=2.09, p=.166, partial h2=.104), nor was the
Time ¥Environment ¥Social Context interaction (F=1.41, p=.251, partial
h2=.072).
Further Characterising the Effects as Restorative
We used the PSS and CFQ to gauge more persistent needs or potential for
restoration at the time of participation in the experiment. On average, the
subjects reported low to moderate levels of perceived stress over the preceding
month at both the first (M=15.10, SD =4.67) and the second (M=15.25,
SD =4.81) administrations of the PSS (scale from 0 to 40). The level of
perceived stress was thus at roughly the same level at the start and then at the
end of the 4 weeks of data collection [paired-samples t(19) =0.17, p=.87]. The
scores from the two PSS administrations correlated strongly, rtest-retest =.65.
Scores on the CFQ indicated that the subjects were sometimes, though not
frequently, troubled by mistakes in perception, memory, and action during the
6 months prior to their participation (M=35.89, SD =7.25) (scale from 0 to
100). Administered only at the start of data collection, before the first walk, the
CFQ scores nonetheless correlated as strongly with the scores from the second
administration of the PSS (rCFQ-PSS2 =.696, p=.001) as they did with the scores
from the first administration (rCFQ-PSS1 =.660, p=.002). We thus combined the
CFQ scores with the two PSS scores into a single measure (a=.84).
Our measure of more persistent restoration needs or potential correlated
strongly and negatively with scores for the positive affects just prior to the
walk, averaged across the four occasions. That is, to the extent that the
subjects indicated more persistent restoration needs, they also reported, just
before their walks, lower levels of Revitalisation (r=-.64, p=.002), Positive
Engagement (r=-.62, p=.004), and Tranquillity (r=-.71, p<.001). They
did not, however, report significantly more Physical Exhaustion (Spearman’s
r=-.07, p=.79), Anxiety/Depression (r=.13, p=.58), Anger (r=-.11,
p=.64), or Time Pressure (r=.19, p=.43). Persistent restoration needs
ENVIRONMENTAL MODERATION OF WALKING BENEFITS 273
© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology.
correlated moderately and negatively, but not significantly, with the mean of
the pre-walk SST scores (r=-.42, p<.07).
Environmental Evaluations
At the end of the experiment the subjects were asked to rank the different
walks in terms of preference (1 =most preferred). The subjects only slightly
preferred the walk in the park with a friend (mean rank =1.90, SD =0.91)
over the walk in the park alone (mean rank =2.05, SD =1.10), which they in
turn only slightly preferred over the walk along streets with a friend (mean
rank =2.35, SD =0.88). They least liked the solitary walk along streets (mean
rank =3.70, SD =0.57). The omnibus test of the difference in rankings
(Friedman’s) yielded c2=24.30 (df =3, N=20, p<.001). Follow-up tests
(Wilcoxon) confirmed what is apparent from the mean ranks; the effect is
owing primarily to dislike for the solitary walk along streets. All three pair-
wise comparisons involving the solitary city walk are significant (psⱕ.001),
but the remaining three comparisons are not significant (ps>.20).
DISCUSSION
We investigated the moderating effects of urban outdoor environment and
immediate social context on changes in affective states and attention expected
to occur while walking. Walking per se did on average reduce Anxiety/
Depression, Anger, and Time Pressure, despite the low-to-moderate initial
levels of those variables. It also increased Revitalisation, Positive Engage-
ment, and Tranquillity. The effect sizes ranged from moderate to large for the
tests of the effects of walking per se (i.e. the main effects of time). This set of
findings concurs with the general result of many studies on immediate emo-
tional benefits of exercise in general (Reed & Ones, 2006) and walking more
specifically (e.g. Ekkekakis et al., 2000, 2008).
A significant effect of walking per se also emerged in the performance of the
symbol-substitution task (SST) that we used to measure the capacity to direct
attention. Contrary to the idea that walking would promote cognitive ben-
efits, however, performance on the SST declined during the walk. Moreover,
it declined to a greater degree with the park walk than with the walk along
city streets. However, the results are inconclusive; the means for the two
environments differed significantly at the pretest but were roughly the same
after the walk, so the change may only reflect regression to the mean.
Environment alone moderated change in feelings of Time Pressure; it
decreased significantly more with the walks in the park than with the walks
along streets. This pattern was mirrored by the results for Tranquillity;
however, the interaction was not statistically significant at the .05 level for that
variable (cf. Bodin & Hartig, 2003).
274 JOHANSSON ET AL.
© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology.
Social context alone moderated change in Physical Exhaustion; on
average, it increased while walking alone, but not while walking with a friend.
Conceivably, this reflects on social facilitation in the form of enjoyable dis-
traction provided by a non-evaluative other in the performance of a simple,
non-competitive activity (cf. Aiello & Douthitt, 2001). Such positive distrac-
tion from the surroundings is also congruent with the idea that the company
of a friend can enhance restoration, though here the outcome reflects less on
restoration than prevention of fatigue.
The urban outdoor environment and the immediate social context together
moderated change in Revitalisation. When walking in the park, Revitalisa-
tion increased more when subjects were alone, whereas when walking along
streets it increased more when subjects were in company. This interaction was
mirrored by the results for Anxiety/Depression; however, for that variable the
three-way interaction was not significant.
The interactive effects of walking, physical environment, and social context
on Revitalisation fit with the results of the scenario study by Staats and
Hartig (2004), which provided a key point of departure for the present
research. In that study, having the company of a friend increased the attrac-
tiveness of a walk in an urban setting, but not of a walk in a forest. As in that
study, subjects in the present study least liked walking along streets alone.
Like the study by Staats and Hartig (2004), this study built on theory and
research concerned with restorative environments in formulating the general
expectation that the environment in which a physical activity takes place
moderates the psychological benefits of that activity. We therefore took
interest in whether our results could be characterised as restorative. Rather
than experimentally inducing a potential for restoration, we used two mea-
sures concerned with persistent deficits in adaptive resources. Given the
low-to-moderate scores on both the PSS and CFQ, and the low-to-moderate
scores for the measures of negative affects just before each walk, it seems
there was rather little need or potential for restoration to begin with on the
different occasions. Yet, we found that to the extent that our subjects
reported persistent resource deficits with the PSS and CFQ, they also
reported less positive affect just before the walks. These associations were
strong. We did not find comparable associations for the negative affects; this
may, however, reflect on the more limited variability in the pre-walk mea-
sures of the negative affects. That aside, any decline in negative affect argu-
ably can be considered as restorative, even when it occurs from an initially
low level.
It is reasonable to ask whether the amount of change in affect and attention
over the course of the walks related to the more persistent potential for
restoration reflected in the combined PSS and CFQ scores. It would,
however, be problematic to simply assess the correlation of restoration needs
with the average amount of change across the occasions. As already shown,
ENVIRONMENTAL MODERATION OF WALKING BENEFITS 275
© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology.
the pre-walk to post-walk change scores were modified by the environment
and/or immediate social context on the walk. Future studies can with more
sophisticated statistical techniques (and the larger samples they require)
address the effects of measured versus induced pre-treatment restoration
needs on the outcomes subsequently realised.
Although the results are consistent with expectations in important respects,
several methodological issues deserve mention. First, we used only single
examples of park and urban street walks. Our limited environmental sam-
pling discourages broad generalisations about the relative benefits of walks in
urban parks and centers, familiar as well as unfamiliar.
Second, with our selection of outdoor environments we unavoidably
selected bundles of social and physical attributes. The street environment
involved not only more exposure to traffic and noise, but also more possi-
bilities for involvement with other people, both known and unknown, than
the park would have provided. Thus, it would be problematic to attribute
beneficial effects to, for example, the amount of greenery alone; parks also
afford relative solitude and less exposure to unwanted sounds from human
activity, for example.
Third, we had a small sample. Despite this, some of the expected
interactions were uncovered. However, we could not confirm the statistical
significance of other interactions, despite considerable effect sizes, as with
the Time ¥Environment interaction in Tranquillity and the Time ¥
Environment ¥Social Context interaction in Anxiety/Depression (cf. Bodin
& Hartig, 2003). The effect sizes reported here can guide power analyses for
future studies and also support meta-analytic work (see Bowler, Buyung-Ali,
Knight, & Pullin, 2010).
Fourth, some of the affect measures had low internal consistency at some
of the test administrations. This would have attenuated associations (Nun-
nally & Bernstein, 1994), and so further compromised the ability to detect
effects already limited by the small sample. In this light it is noteworthy that
we could uncover significant effects despite the sometimes low internal con-
sistency, particularly with the Revitalisation measure.
Fifth, we did not obtain measures of affect during the walk. Conceivably,
our subjects experienced variations in affects during the walk that are not
reflected in the difference between the pre-walk and post-walk measures (e.g.
increase and then dissipation by the end; cf. Hull, Michael, Walker, &
Roggenbuck, 1996). As it stands, however, we did capture significant change
on average for all but one of the affect measures.
Finally, the strong main effects of walking per se on affective states may to
some extent reflect the operation of demand characteristics; the subjects may
have anticipated that they were expected to show improvements from pre-
walk to post-walk. Such an explanation seems less plausible, however, with
regard to the various interaction effects.
276 JOHANSSON ET AL.
© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology.
In sum, we provide evidence that some affective benefits of a brisk walk
depend in varying ways on the outdoor urban environment and the immedi-
ate social context. Although small in scale, this experiment fruitfully joins and
builds on two established lines of research, one concerned with the restorative
effects of different environments (e.g. Hartig, Böök, Garvill, Olsson, &
Gärling, 1996; Hartig et al., 1991, 2003) and the other with the manner in
which restoration needs and characteristics of the socio-physical environment
together affect preferences regarding environments in which restoration
might be sought (e.g. Staats, Kieviet, & Hartig, 2003: Staats & Hartig, 2004).
Further research of this kind can follow up on the present findings and the
findings of Staats et al. (2010) by comparing the effects of a broader range of
solitary and social activities-in-environments that urban residents value for
restoration. Such work addresses an important gap in theorising about restor-
ative environments. Theory and research to date have identified social causes
and consequences of different forms of resource depletion (Kaplan, 1995),
and so have indicated potential social benefits of restorative processes:
however, little work has yet directly addressed specific social variables that
sustain restorative processes in different environments (Staats & Hartig,
2004). We think our propositions concerning the role of company in enabling
and enhancing restoration constitute a useful new contribution in this respect.
Results of this kind of research suggest possible consequences of urban
design and planning alternatives that open or close various opportunities for
physical activity. They further affirm the notion that having a companion can
offset some of the deficits in the broader socio-physical environment. Still,
while the company of a friend may enhance a walk along streets in terms of
some psychological benefits, our results also suggest that other potential
benefits are more dependent on the broader socio-physical environment,
regardless of the immediate social context. In particular, together with the
results of other field experiments in this area (e.g. Hartig et al., 1991, 2003;
Bodin & Hartig, 2003; Berman et al., 2008), some of the present results speak
to the value of what Pretty and colleagues (2005, 2007) have called “green
exercise”; that is, exercise in relatively natural environments like parks. For
the realisation of some psychological benefits of exercise, and for the reali-
sation of benefits with physical activity more generally, it appears that parks
and other settings that afford contact with nature offer a distinct advantage
for urban residents.
ACKNOWLEDGEMENTS
We thank Tora Holmberg, Per Gustafson, Anders Lindbom, Helen Ekstam,
Anthony Ong and the anonymous reviewers for their constructive comments
on an earlier draft of this paper.
ENVIRONMENTAL MODERATION OF WALKING BENEFITS 277
© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology.
REFERENCES
Aiello, J.R., & Douthitt, E.A. (2001). Social facilitation from Triplett to electronic
performance monitoring. Group Dynamics: Theory, Research, and Practice,5,
163–180.
Berman, M.G., Jonides, J., & Kaplan, S. (2008). The cognitive benefits of interacting
with nature. Psychological Science,19, 1207–1212.
Bodin, M., & Hartig, T. (2003). Does the outdoor environment matter for psycho-
logical restoration gained through running? Psychology of Sport and Exercise,4,
141–153.
Bowler, D.E., Buyung-Ali, L.M., Knight, T.M., & Pullin, A.S. (2010). A systematic
review of evidence for the added benefits to health of exposure to natural environ-
ments. BMC Public Health,10, 456.
Broadbent, D.E., Cooper, P.F., FitzGerald, P., & Parkes, K.R. (1982). The Cognitive
Failures Questionnaire (CFQ) and its correlates. British Journal of Clinical Psy-
chology,21, 1–16.
Cohen, S. (1978). Environmental load and allocation of attention. In A. Baum, J.E.
Singer, & S. Valins (Eds.), Advances in environmental psychology (Vol. 1, pp. 1–29).
Hillsdale, NJ: Erlbaum.
Cohen, S., & Williamson, G. (1988). Perceived stress in a probability sample of the
United States. In S. Spacapan & S. Oskamp (Eds.), The social psychology of health
(pp. 31–67). Newbury Park, CA: Sage.
De Vries, S., Classen, T., Hug, S.M., Korpela, K., Maas, J., Mitchell, R. et al.
(2011). Contributions of natural environments to physical activity: Theory and
evidence base. In K. Nilsson, M. Sangster, C. Gallis, T. Hartig, S. De Vries, K.
Seeland et al. (Eds.), Forests, trees, and human health (pp. 205–243). Dordrecht:
Springer.
Ekkekakis, P., Backhouse, S.H., Gray, C., & Lind, E. (2008). Walking is popular
among adults, but is it pleasant? A framework for clarifying the link between
walking and affect as illustrated in two studies. Psychology of Sport and Exercise,
9, 246–264.
Ekkekakis, P., Hall, E.E., van Landuyt, L.M., & Petruzello, S.J. (2000). Walking in
(affective) circles: Can short walks enhance affect? Journal of Behavioral Medicine,
23, 245–275.
Evans, G.W., & Cohen, S. (1987). Environmental stress. In D. Stokols & I. Altman
(Eds.), Handbook of environmental psychology: Vol. 1 (pp. 571–610). New York:
Wiley.
Frumkin, H., Frank, L., & Jackson, R. (2004). Urban sprawl and public health:
Designing, planning, and building for healthy communities. Covello, CA: Island
Press.
Gauvin, L., & Rejeski, W.J. (1993). The Exercise-Induced Feeling Inventory: Devel-
opment and initial validation. Journal of Sport and Exercise Psychology,15, 403–
423.
Hartig, T., Böök, A., Garvill, J., Olsson, T., & Gärling, T. (1996). Environmental
influences on psychological restoration. Scandinavian Journal of Psychology,37,
378–393.
278 JOHANSSON ET AL.
© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology.
Hartig, T., Evans, G.W., Jamner, L.D., Davis, D., & Gärling, T. (2003). Tracking
restoration in natural and urban field settings. Journal of Environmental Psychol-
ogy,23, 109–123.
Hartig, T., Korpela, K., Evans, G.W., & Gärling, T. (1997). A measure of restorative
quality in environments. Scandinavian Housing and Planning Research,14, 175–194.
Hartig, T., Mang, M., & Evans, G.W. (1991). Restorative effects of natural environ-
ment experience. Environment and Behavior,23, 3–26.
Haskell, W.L., Lee, I.-M., Pate, R.P., Powell, K.E., Blair, S.N., Franklin, B.A. et al.
(2007). Physical activity and public health: Updated recommendation for adults
from the American College of Sports Medicine and the American Heart Associa-
tion. Circulation,116, 1081–1093.
Hassmén, P. (1996). Environmental effects on ratings of perceived exertion in males
and females. Journal of Sport Behavior,19, 235–245.
Herzog, T.R., Maguire, C.P., & Nebel, M.B. (2003). Assessing the restorative com-
ponents of environments. Journal of Environmental Psychology,23, 159–170.
Hug, S.-M., Hansmann, R., Monn, C., Krütli, P., & Seeland, K. (2008). Restorative
effects of physical activity in forests and indoor settings. International Journal of
Fitness,4, 25–38.
Hull, R.B., Michael, S.E., Walker, G.J., & Roggenbuck, J.W. (1996). Ebb and flow of
brief leisure experiences. Leisure Sciences,18, 299–314.
Kaplan, R., & Kaplan, S. (1989). The experience of nature: A psychological perspec-
tive. Cambridge: Cambridge University Press.
Kaplan, S. (1995). The restorative benefits of nature: Toward an integrative frame-
work. Journal of Environmental Psychology,15, 169–182.
Kerr, J.H., Fujiyama, H., Sugano, A., Okamura, T., Chang, M., & Onouha, F. (2006).
Psychological responses to exercising in laboratory and natural environments.
Psychology of Sport and Exercise,7, 345–359.
Laumann, K., Gärling, T., & Stormark, K.M. (2001). Rating scale measures of
restorative components of environments. Journal of Environmental Psychology,21,
31–44.
Laumann, K., Gärling, T., & Stormark, K.M. (2003). Selective attention and heart
rate responses to natural and urban environments. Journal of Environmental
Psychology,23, 125–134.
Lezak, M.D. (1983). Neuropsychological assessment (2nd edn.). Oxford: Oxford
University Press.
McAuley, E., Blissmer, B., Katula, J., & Duncan, T.E. (2000). Exercise environment,
self-efficacy, and affective responses to acute exercise in older adults. Psychology
and Health,15, 341–355.
Martinsen, E.W., & Raglin, J.S. (2007). Anxiety/depression: Lifestyle medicine
approaches. American Journal of Lifestyle Medicine,1, 159–166.
Milgram, S. (1970). The experience of living in cities. Science,167, 1461–1468.
Nunnally, J.C., & Bernstein, I.H. (1994). Psychometric theory (3rd edn.). New York:
McGraw-Hill.
Plante, T.G., Coscarelli, L., & Ford, M. (2001). Does exercising with another enhance
the stress-reducing benefits of exercise? International Journal of Stress Manage-
ment,8, 201–213.
ENVIRONMENTAL MODERATION OF WALKING BENEFITS 279
© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology.
Pretty, J., Peacock, J., Hines, R., Sellens, M., South, N., & Griffin, M. (2007). Green
exercise in the UK countryside: Effects on health and psychological well-being,
and implications for policy and planning. Journal of Environmental Planning and
Management,50, 211–231.
Pretty, J., Peacock, J., Sellens, M., & Griffin, M. (2005). The mental and physical
health outcomes of green exercise. International Journal of Environmental Health
Research,15, 319–337.
Reed, J., & Ones, D.S. (2006). The effect of acute aerobic exercise on positive acti-
vated affect: A meta-analysis. Psychology of Sport and Exercise,7, 477–514.
Scott, B., Brandberg, M., & Öhman, A. (2001). Measuring the negative mood com-
ponent of stress experiences: Description and psychometric properties of a short
adjective check-list of stress responses. Journal of Psychology,42, 1–7.
Staats, H., & Hartig, T. (2004). Alone or with a friend: A social context for psycho-
logical restoration and environmental preferences. Journal of Environmental Psy-
chology,24, 199–211.
Staats, H., Kieviet, A., & Hartig, T. (2003). Where to recover from attentional fatigue:
An expectancy-value analysis of environmental preference. Journal of Environmen-
tal Psychology,23, 147–157.
Staats, H., van Gemerden, E., & Hartig, T. (2010). Preference for restorative situa-
tions: Interactive effects of attentional state, activity-in-environment and social
context. Leisure Sciences,32, 401–417.
Tabachnik, B.G., & Fidell, L.S. (2001). Using multivariate statistics (4th edn.).
Needham Heights, MA: Allyn and Bacon.
Ulrich, R.S., Simons, R., Losito, B.D., Fiorito, E., Miles, M.A., & Zelson, M. (1991).
Stress recovery during exposure to natural and urban environments. Journal of
Environmental Psychology,11, 201–230.
United Nations (2002). World urbanization prospects: The 2001 revision. Geneva: UN
Population Division.
United States Department of Health and Human Services (USDHHS) (1996). Physi-
cal activity and health: A report of the Surgeon General. Atlanta, GA: US Depart-
ment of Health and Human Services, Centers for Disease Control and Prevention,
National Center for Chronic Disease Prevention and Health Promotion.
Wendel-Vos, W., Droomers, M., Kremers, S., Brug, J., & van Lenthe, F. (2007).
Potential environmental determinants of physical activity in adults: A systematic
review. Obesity Reviews,8, 425–440.
280 JOHANSSON ET AL.
© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology.
