Environment and Behavior
© The Author(s) 2018
Article reuse guidelines:
Health Benefits of
Walking in Nature: A
Study Under Conditions
of Real-Life Stress
Gunnthora Olafsdottir1,2,3 , Paul Cloke2,
André Schulz1, Zoé van Dyck1,
Thor Eysteinsson4, Björg Thorleifsdottir4,
and Claus Vögele1
We investigated the effects of recreational exposure to the natural
environment on mood and psychophysiological responses to stress.
We hypothesized that walking in nature has restorative effects over and
above the effects of exposure to nature scenes (viewing nature on TV) or
physical exercise alone (walking on a treadmill in a gym) and that these
effects are greater when participants were expected to be more stressed.
Healthy university students (N = 90) were randomly allocated to one of
three conditions and tested during an exam-free period and again during
their exam time. Mood and psychophysiological responses were assessed
before and after the interventions, and again after a laboratory stressor.
All interventions had restorative effects on cortisol levels (p < .001), yet
walking in nature resulted in lower cortisol levels than did nature viewing
1University of Luxembourg, Esch-sur-Alzette, Luxembourg
2University of Exeter, Exeter, U.K.
3Icelandic Tourist Board, Reykjavík, Iceland
4University of Iceland, Reykjavik, Iceland
Gunnthora Olafsdottir, Icelandic Tourist Board, Geirsgata 9, 101 Reykjavík, Iceland.
800798EABXXX10.1177/0013916518800798Environment and BehaviorOlafsdottir et al.
2 Environment and Behavior 00(0)
(p < .05) during the exam period. Walking in nature improved mood more
than watching nature scenes (p < .001) or physical exercise alone (p < .05).
restorative environments, natural environments, stress, walking, mental and
physical health, psychophysiology, psychological wellbeing, environmental
Through the centuries, human beings have been drawn to nature to relax,
recuperate, and find temporary freedom from the stressors of everyday life
(Herzog, Black, Fountaine, & Knotts, 1997; Home, Hunziker, & Bauer, 2012;
Knopf, 1987). These ventures are based on the age-old belief in the salubri-
ous effects of exposure to nature (Parsons, 1991; Ulrich, 1979). Stress relief,
escaping from civilization, clearing the head, reflecting on important life
issues, experiencing beauty, and connecting with nature are among the domi-
nant self-reported motives (Driver, 1976; Knopf, 1987). The single most
important self-reported benefit of exposure to both nature-rich urban places
and wilderness areas, however, is stress mitigation (Ulrich et al., 1991).
Investigations of the causal mechanisms underlying such mediated health
effects of the natural environment are apposite, significant, and pressing,
especially in the light of the fact that stress is generally recognized as the
most severe health problem in the 21st century (Kudielka & Wüst, 2010;
Tyrväinen et al., 2014).
Restorative effects of nature exposure have been psychologically and
physiologically accounted for in terms of “reduction in cognitive fatigue,
decreased stress levels, increased focus, increased positive affect,
decreased negative affect, and decreased sympathetic nervous system
activity” (Valtchanov, Barton, & Ellard, 2010, p. 503). These results are
interpreted in the context of the two dominating theories of nature restora-
tion: (a) attention restoration theory (ART) that focuses on the detrimental
factors of mental fatigue on our capacity to direct attention and associates
the restorative effect of being in nature with its potential to conjure up soft
fascination, which is the remedy needed to rest and recover (S. Kaplan,
2001; R. Kaplan & Kaplan, 1989; S. Kaplan & Talbot, 1983) and (b) the
psychophysiological stress reduction framework, which posits an immedi-
ate positive affective and stress-releasing response to nature as a non-
threatening environment (Ulrich, 1981, 1983, 1984; Ulrich et al., 1991).
Both theories are based in part on the assumption of the innate affiliation
Olafsdottir et al. 3
to nature, and presume that nature can promote faster and more complete
restoration than other environments for mentally fatigued and stressed
individuals (Hartig, Mitchell, de Vries, & Frumkin, 2014; R. Kaplan &
Kaplan, 1989; Ulrich, 1981).
Correlational and experimental findings over the past 30 years provide
evidence for these restorative effects of nature (Health Council of the
Netherlands, 2004; James, Banay, Hart, & Laden, 2015; Kondo, Fluehr,
McKeon, & Branas, 2018). In the context of exercise, some experimental
findings based on evidence at psychological (Pretty, Griffin, Sellens, &
Pretty, 2003), physiological (Gladwell et al., 2012; Park, Tsunetsugu,
Kasarani, Kagawa, & Miyazaki, 2010), and biochemical (Li et al., 2011)
levels indicate that exercise with nature exposure (i.e., green exercise) is
more beneficial than the same exercise in built-up places (Pretty et al.,
2003). Nevertheless, a recent review shows that the evidence linking
restorative physiological changes in response to nature exposure via exer-
cise (walking or running) in comparison to other environments is sparse
(Gladwell, Brown, Wood, Sandercock, & Barton, 2013). Reviews of the
health benefits of passive and active activities in nature versus indoor,
built (and synthetic) urban environments conclude that—in comparison—
nature is associated with improved mental health and well-being (Bowler,
Buyung-Ali, Knight, & Pullin, 2010; Ohly et al., 2016; Thompson-Coon
et al., 2011). In a systematic review, including studies using several physi-
ological parameters (blood pressure [BP], heart rate [HR], heart rate vari-
ability [HRV], cortisol, noradrenaline, and adrenaline; Bowler et al.,
2010), there was not enough evidence to suggest a differential effect in
favor of the natural environment due to a lack of studies of comparable
design (Bowler et al., 2010; see also the review by Hartig et al., 2014).
Nevertheless, some studies suggest differences in the effects of nature ver-
sus urban environments on physiological stress responses (Brown, Barton,
& Gladwell, 2013; Gladwell et al., 2012; Gladwell et al., 2013; Laumann,
Garling, & Stormark, 2003; Li et al., 2011; Park et al., 2010; Pretty,
Peacock, Sellens, & Griffin, 2006; Ulrich, 1981; Ulrich et al., 1991).
Brown and colleagues (2013), for example, investigated the effects of
viewing images containing nature scenes compared with scenes from the
built environment before exposure to a laboratory mental stressor. Although
cardiovascular responses to the stressor did not differ between conditions,
cardiac-vagal activity as indexed by HRV was higher during recovery in
the nature-scene viewing condition.
Although such results are promising, it appears that the current evidence
is often limited by the use of picture viewing (Brown et al., 2013; Gladwell
et al., 2012; Laumann et al., 2003; Parsons, 1991; Pretty et al., 2006; Ulrich,
4 Environment and Behavior 00(0)
1981; Ulrich et al., 1991) rather than real-life situations, thereby lacking
ecological validity. It has been argued that using passive viewing of images
as a condition rather than real exposure to nature allows for the exclusion
of any confounding effects of exercise, which is often part of nature experi-
ence and has itself inherent health benefits (Brown et al., 2013).
Nevertheless, it seems important to use a research design that allows for the
systematic investigation of the differential effects of all the components
involved in nature exposure in a real-life setting. The first aim of the current
study, therefore, was to compare the effects of nature exposure on mood
and psychophysiological stress responses, and to disentangle the effects
associated with recreational nature exposure from those linked with exer-
cise (walking) by comparing three conditions: walking in nature, watching
a video of the same nature scenes, and walking on a treadmill in a gym.
Another shortcoming of the current literature on the effects of nature
exposure on stress and well-being is the predominant use of laboratory stress-
ors. The specific purpose of laboratory stress paradigms is to elicit psycho-
physiological responses to emotional or behavioral challenges independent
of basic reflexes so as to evaluate the role of the central nervous system in
disturbances of stress regulation and the involvement of stress mechanisms in
physical and mental disorders. There are several features of mental stress
testing in the laboratory that make it an attractive research paradigm, includ-
ing ease of application, high degree of standardization, and the possibility to
monitor psychophysiological stress responses using sophisticated physiolog-
ical measurements (Steptoe & Vögele, 1991). Nevertheless, in daily life peo-
ple rarely confront a single stressor and usually face a myriad of stressors
from many sources and of varying duration. Especially chronic stress forces
a person to continually adapt to changing circumstances in the medium- and
long-term, although “chronic” is a poorly defined construct. In a meta-analy-
sis investigating the link between acute stress responsivity to laboratory
stressors and changes in various chronic psychosocial stress conditions (e.g.,
job stress, general life stress), there were no associations between chronic
psychosocial stress and stress reactivity or recovery as assessed by parame-
ters of sympathetic and parasympathetic nervous system activity (Chida &
Steptoe, 2009). The use of laboratory stressors in lieu of real-life stress is,
A second aim of the current study was, therefore, to extend the current
literature on the potential stress-buffering effects of nature exposure by
including a more chronic life stress (academic examination period), in
addition to a standardized laboratory stressor. Participants attended the
psychophysiological experiment twice, that is, under academic stress
(exam period) and no-stress conditions (no exam period). This repeated
Olafsdottir et al. 5
measurement design allowed for the investigation of differential and inter-
action effects of both types of stressors on psychophysiological reactivity
in relation to the three treatment conditions. We hypothesized that (a)
exposure to nature in the context of leisure walking has restorative effects
on mood and psychophysiological stress responses under chronic and
acute stress conditions over and above the effects of passive exposure to
nature scenes or physical exercise alone, and (b) that the restorative effects
of nature exposure would be more pronounced during the exam period
when the participants were expected to be more stressed. We did not expect
to find differences between exercise and passive viewing of nature scenes
as both interventions have been shown to be effective.
Sample and Research Design
The study used a mixed factorial design. It was carried out in Reykjavik
with prior ethics approval from the Ethics Review Panel of the University
of Luxembourg and the National Bioethics Committee in Iceland. Overall,
90 volunteers were recruited from the local universities. Advertisements
describing the study were sent by email to all students via student unions.
Permission was sought from the university authorities to advertise on uni-
versity websites and to put printed copies on advertisement boards. The
same advertisement was posted in supermarkets and on student campuses
and (a small version) published in a local newspaper. The advertisement
invited students to contact the Principal Investigator (PI) for more infor-
mation. Eligibility for participation was assessed via telephone screening
of study volunteers. Eligible participants were randomly allocated to one
of three 40-min (±5 min) activities: (a) walking in nature, representing
direct exposure to nature (nature), (b) walking (on a treadmill) in a gym
representing the same physical activity intensity without nature exposure
(gym), and (c) watching nature video-recording on TV in a laboratory set-
ting (video), representing exposure to the same natural environment with-
out the physical effort of walking (Figure 1). The sample size was based on
power analysis, assuming an alpha level of significance of 5%, and test
power above 90% (91.88%). Exclusion criteria were as follows: (a) self-
reported current and regular exercise in nature or non-nature environ-
ments, (b) self-reported current ill health and/or diagnosed chronic
conditions that preclude participation in physical exercise and/or a stress
test, including substance abuse, and (c) skin problems and known allergic
reactions to electrodes. Data were collected in two waves: from January to
6 Environment and Behavior 00(0)
March 2014, that is, a period when students were not taking exams, and
again from April to May 2014 during the exam period. To ensure that this
operationalization of chronic life stress was valid, participants were asked
to provide information on the date of their next examination. There were
no significant differences in mean days until the next exam between the
Figure 1. Consort flowchart.
Note. HR = heart rate; HRV = heart rate variability.
Olafsdottir et al. 7
groups. Individual experimental sessions took place in the afternoon for
reasons of standardization concerning the circadian rhythm of cortisol
secretion and other dependent variables. Six participants dropped out
before the second measurement. Their data along with responses from 12
participants who did not complete all measurements and five others with
health-related issues not known during recruitment (pregnancy, depres-
sion, heart condition) was excluded from the analysis. The final sample
consisted of 67 eligible participants (age 20-33 years; M = 24.39 years, SD
= 2.61 years) with 20 participants in the nature group (15 females), 24 in
the gym group (16 females), and 23 in the video group (15 females). The
test power for the final sample was 83%. Mean body mass index (BMI) for
the gym group was 22.37 kg/m2 (SD = 2.96), 24.22 kg/m2 for the nature
group (SD = 2.96), and 25.19 kg/m2 for the video group (SD = 4.56).
There were no significant differences in BMI between the groups.
The nature walk took place in a conserved and by far the largest recreational
area of Reykjavík city (Figure 2a). The majority of the 32 km2 are covered
Figure 2. (a) the nature-setting in May (upper left), (b) the nature-setting in
February (upper right), (c) the gym-setting (lower left), (d) the video-setting
Source. Author (All figures in color on-line only).
8 Environment and Behavior 00(0)
with woodland of 26 different species predominated with the Sitka Spruce
(Picea sitchensis). The experiment took place on a 4.08 km sign-posted and
easy circle trail (elevation 37 m) through the woodland. From time to time,
during the walk the environment surrounding the trail opens up to open
green spaces, moss-covered lava, and surrounding mountains. Mid-way,
after 2.13 km into the trail, the outskirts of the city, Faxafloi bay and the
mountains can be seen in the far distance. For the latter half, from 2.3 km
onwards, the environment is again limited to greenery, lava, and the sky. We
selected this trail, as people tend to like easily accessible environments char-
acterized by trees and water (Ulrich, 1979, 1986) that can be walked in a
relatively short period of time (40 ± 5 min). Compared with previous stud-
ies (Li et al., 2011; Park et al., 2010), we chose a trail with a longer walking
time to improve ecological validity and also because the length of the walk
could be easily matched in a gym setting without being out of place.
Moreover, the route was also chosen as the only sign-posted route within the
recreational area that does not involve crossing the main road, and starts and
ends at a parking lot where the study’s mobile laboratory could be placed.
This way the laboratory marked the start and finish of the walk and the psy-
chophysiological measurements could be done immediately before and after
the individual walks in nature. As the experiment took place in the last
months of the Icelandic winter (February-March) and again at the beginning
of spring (April-May), all participants in the outdoor setting encountered
snow during the first period, but none during the second (Figure 2b). In
February/March, the average temperature, wind factor, and cloud coverage
was higher than in April/May (5.1°C vs. 4.6°C; 6.3 m/sec vs. 4.2 m/sec; 5.6
of 8 vs. 7 of 8 parts, respectively), but precipitation and hours of sunshine
were lower (0.7 mm vs. 1.6 mm; 0.7 hr vs. 8.5 hr of sunshine).
The indoor walking exercise took place at a gym selected because of its
easily accessible location, modern man-made infrastructure, and its goal of
promoting exercise for health and well-being; this compared better to the
nature walk intervention than a conventional body-building gym (Figure 2c).
All participants in the gym group used the same treadmill (Cybex 750T) in
the main equipment room, surrounded by other equipment and customers.
For participants in both the gym and the nature walk group, there was a pos-
sibility of encountering other people during their respective 40-min walk.
These are ecological conditions normal to these settings in the early after-
noon. Participants were instructed not to engage in any interaction with oth-
ers during their 40-min walk.
The video nature viewing took place under laboratory conditions at the
University of Iceland (Figure 2d). A small room within the laboratory was
furnished with a comfortable armchair that was situated in front of a 55″
Olafsdottir et al. 9
color TV set. Following the recommendations of the Society for Motion
Picture and Television Engineers (SMPTE), we chose a viewing angle of 43°
(recommended range 61.8° to 33.3°; SMPTE standard EG-18-1994).
Accordingly, the chair for viewers was placed 128 cm from the screen.
The video recording was obtained from the same trail the participants
allocated to the nature group walked. The recording took place in
December, and again in March, 2 weeks prior to each viewing so that the
sights and sounds of nature matched the season and to ensure similar expo-
sure to nature as the nature group. For the first shooting in December, the
footage consisted of a simulated walk with an adapted steadicam-like
mechanism and approximately 2 min x 40 min in AVCHD 25M bit/sec in
1920 x 1080 50i shot on a Sony CX730. In interesting places, static,
detailed, and panoramic shots were included to simulate a rest (e.g., on a
bench). The simulated walk resulted in a lot of perturbations due to icy,
slippery parts of the path and walking movement while recording the
video. We, therefore, recorded a second video with a SONY FS700
NXCAM Codec and 1920 x 1080 50i to obtain more high quality pictures
with a fixed camera position (panoramic movement, details, or panora-
mas). This shooting also generated a video of approximately twice the
duration of the walk (80 min). Sound setup included two crossed Sennheiser
shotgun microphones in 90° XY setup to get the best possible stereo
recording for the sound ambience.
The settings for the video recording in March were the same as in
December, that is, a Sony FS700 NXCAM, however, with a modified sound
setup including a Rode NT1000 with Sennheiser shotgun in MS-Stereo
recording. This resulted in 100 min of footage including details, panoramas,
and walks to simulate the movement between the different points of interest.
The footage was taken in 25 fps and played back in 100% speed. Some spe-
cial settings and panoramas (e.g., moving clouds) were shot in time-lapse
(10x) to better capture the image of the moving clouds on the beautiful land-
scape. Both video footages (i.e., winter and spring) were edited to each match
a 40-min walk, resulting in two 40-min video clips.
The Socially Evaluated Cold-Pressor Test (SECPT; Schwabe, Haddad, &
Schächinger, 2008) was used to induce acute stress. This is a well-established
and standardized method eliciting significant cardiovascular (HR) and endo-
crinological stress responses (e.g., cortisol). The test involves immersing the
nondominant hand and wrist in cold water (0ºC-3ºC) for 3 min while being
monitored and videotaped for the mock purpose of later analysis of facial
10 Environment and Behavior 00(0)
expressions. The socially evaluative component (experimenter, fake video
recording) is required to provoke cortisol responses (Schwabe, Haddad, &
Schächinger, 2008). Its focused experimental setup allows for a precise tim-
ing of evoked physiological stress responses.
Cortisol assays. To assess hypothalamic pituitary adrenocortial (HPA) axis
activity, cortisol was sampled and assayed from participants’ saliva. Saliva
samples were collected using salivettes (Sarstedt ref: 51.1534.500) and fro-
zen at −20°C immediately after the sessions. Salivary samples were assayed
twice at the Department of Clinical Biochemistry at the National University
Hospital of Iceland, with the highly standardized nonisotopic and automatic
electrochemiluminescence sandwich immunoassay (Mathew, Biju, & Thapa-
lia, 2005) using kits provided by Roche Diagnostic GmbH, Mannheim, Ger-
Cardiac data. HR and HRV were derived from the monitored electrocar-
diogram (ECG). For this measurement, we followed the standard method-
ological procedures (Task Force of the European Society of Cardiology
and the North American Society of Pacing Electrophysiology [TFESC and
NASPE], 1996). ECG was monitored with three electrodes (two active and
one ground) placed on the individual’s chest using a single standard lead I
(Einthoven) configuration. ECG was recorded with a NeXus-32 biofeed-
back system with a sampling rate of 1,024 Hz. The spectral analysis of the
manually confirmed RR-interval series was carried out using a Fast Fou-
rier Transformation (FFT; 5 Hz resampling, Hanning window, triangular
weighting). We followed the recommendations by the TFESC and NASPE
and defined the high frequency (HF) band as 0.15 Hz to 0.4 Hz, and the
low frequency (LF) band as 0.04 Hz to 0.15 Hz. HRV was expressed in
terms of HFnu (normalized power in the high-frequency band [0.15 Hz-0.4
Hz]) as time-domain derived HRV parameters (e.g., Root Mean Square of
the Successive Differences [RMSSD]) would have been inappropriate as
trials differed in duration.
Mood was assessed using the Positive and Negative Affect Scale, a self-
report questionnaire that consists of two 10-item scales to measure both posi-
tive and negative affect (Positive and Negative Affect Schedule [PANAS];
Crawford & Henry, 2004; Watson, Clark, & Tellegen, 1988). In this
Olafsdottir et al. 11
questionnaire, participants evaluate 20 mood-related adjectives according to
their current perceived mood and rate them on a 5-point scale from 1 = very
slightly or not at all to 5 = extremely. Questions 1, 3, 5, 9, 11, 12, 14, 16, 17,
19 refer to positive affect (PA) whereas 2, 4, 6, 7, 8, 10, 13, 15, 18, 20 refer
to negative affect (NA). The score of the negative items are reversed in the
analysis, and the total NA state score is found by adding up the sum of the NA
items and PA items for PA. The total state score for each can range from 10 to
50. Higher scores indicate higher levels of NA and PA. Mean scores for PA is
33.3 (SD ± 7.2) and 17.4 (SD ± 6.2) for NA (Watson et al., 1988). The ques-
tionnaire was administered three times during the trial (quantified as NA and
PA). The PANAS is a widely used instrument to reflect perceived stress, as its
NA state scale is sensitive for state anxiety, which represents one important
affective component of acute stress perception (Rossi & Pourtois, 2012).
Furthermore, the PANAS allows for the assessment of possible effects of
nature walking on PA.
A research diary was kept to record qualitative information on environmental
conditions of each experimental trial, recording factors of the weather (nature
group) that was substantiated with the quantitative measures of the Icelandic
Met Office. Six randomly selected participants from each of the three groups
were interviewed after the experimental sessions about their experience of
the interventions. Results of these qualitative data are reported elsewhere
(Olafsdottir, Cloke, & Vögele, 2017).
Participants were instructed not to arrive hungry or thirsty at the study set-
ting, and to refrain from smoking, eating, and drinking 1 hr prior to their
individual afternoon appointments. The conditions in situ were kept as con-
trolled as possible. On assessment days, the PI walked the nature route 1 hr
prior to the experiment to ensure that conditions were safe and suitable for
participants. The order and sequence of measurements within each trial was
always the same (Figure 3). Before the intervention, and after receiving
informed consent, ECG was recorded under resting conditions for 10 min.
Thereafter, saliva was sampled for cortisol and the PANAS administered.
Thereafter, the study participant performed the intervention alone. For the
nature walk, participants were introduced to the route in the park’s user-
friendly trail map. They were thoroughly informed about the route’s condi-
tions (e.g., elevations, slippery in places in winter), the signposts and
12 Environment and Behavior 00(0)
instructed to keep a comfortable walking speed, not a speed that makes one
sweat, but allows one to enjoy the walk and the setting. Participants were
informed that in good weather conditions the route could be walked in 40
min ± 5 min. More clothes (parka, hat, mittens) were offered as needed, to
ensure comfort en route.
In the gym, participants were guided to the equipment hall. After familiar-
ization with the equipment, the clock on the treadmill was visibly set to 40 min.
Volunteers were asked to step on the treadmill, switch it on and select a speed
for walking, similar to the walking speed used on a leisurely walk outside. The
speed was supposed to be comfortable walking speed for the individual that
would not break a sweat. It could be adjusted by the participant whenever
needed but kept within the range of 2.6 km/hr to 4.6 km/hr, which accommo-
dated different levels of fitness and stride length.
Participants in the video group were invited to enter a small room within the
laboratory and sit down in front of the TV. Before the video was switched on,
participants were instructed to watch the video from start to finish and then re-
enter the laboratory for measurements.
Post-intervention measurements started immediately after the intervention.
Cortisol was sampled while the participant was reconnected with the ECG
device for a 5-min recording period. The PANAS was then repeated. Next the
stressor (SECPT) was introduced. A pre-adjusted video camera was brought into
the room along with a bucket of water. The video recording started and partici-
pants were instructed to immerse their nondominant hand up to and including
the wrist in the cold-water bath for 3 min and face the camera, but allowed to
remove it prematurely if feeling very uncomfortable. Then, the video camera
was switched off, and participants were handed a towel to dry their hand and to
Figure 3. The trial sequence of the study.
Note. ECG = electrocardiogram; C = saliva collection for cortisol analysis; M = mood;
t1–t4 = ECG measurement points.
Olafsdottir et al. 13
speed recovery. The PANAS was then answered for the third time. ECG was
recorded throughout the procedure and for 10 min after the stressor. The third
saliva sample was collected 18 min after removing the hand from the water, as
previous studies on SECPT effects on cortisol responses indicated the peak of
cortisol release between 15 and 20 min after SECPT termination (Lass-
Hennemann et al., 2010; Schwabe, Haddad, & Schächinger, 2008). Participants
were then debriefed and given time to recover after the procedure (Figure 3).
To evaluate the effects of exposure to nature on cortisol levels under chronic and
acute stress conditions, a 3 x 3 x 2 mixed-design analysis of variance (ANOVA)
was conducted, with Group (nature, video, gym) as the between-subject factor,
and Time (baseline, after intervention, after SECPT) and Period (no exams,
exams) as the within-subject factors. The effects of exposure to nature on cardio-
vascular data were analyzed using two separate 3 x 4 x 2 mixed-design ANOVAs,
with Group (nature, video, gym) as the between-subject factor, Time (baseline
[t1], after intervention [t2], during SECPT [t3], follow-up [t4]) and Period (no
exams, exams) as the within-subject factors, and with HR and HFnu as depen-
dent variables. Because of equipment failure, ECG data of 17 participants had to
be excluded from analyses, resulting in a final sample of 50 participants for
ECG analyses (nature: n = 17; gym: n = 22; video: n = 11). To assess the
effects of exposure to the natural environment on psychological measures, we
calculated two separate 3 x 3 x 2 mixed-design ANOVAs, with Group (nature,
video, gym) as the between-subject factor, Time (baseline, after intervention,
after SECPT) and Period (exams, no exams) as the within-subject factors, and
with NA and PA as dependent variables. Significant effects and interactions
were assessed with additional t tests as post hoc comparisons with Bonferroni or
Games-Howell corrections, depending on homogeneity of variance, as deter-
mined by Levene’s test for equality of variances. Effects of potential baseline
differences were controlled for by introducing baseline levels as covariates. In
these cases, the initial 3 x 3 x 2 ANOVA model for cortisol and psychological
measures was modified into a 3 x 2 x 2 analysis of covariance (ANCOVA),
whereas the 3 x 4 x 2 ANOVA for cardiovascular data was changed into a 3 x 3
x 2 ANCOVA (identical factors, excluding “baseline” from factor “time”). All
results are expressed as mean ± standard deviation. A value of p < .05 was
considered statistically significant. Statistical software SPSS version 22.0 (SPSS
Inc., Chicago, IL, USA) was used for statistical analysis.
Preliminary analyses using sex as a grouping factor revealed no signifi-
cant differences between men and women in any of the dependent variables.
Results are, therefore, reported for both men and women together.
14 Environment and Behavior 00(0)
Cortisol. Analysis of cortisol levels showed significant main effects for Period
F(1, 61) = 32.91, p < .001, η2 = .35 and Time F(1, 122) = 36.41, p < .001, η2
= .37, together with a significant Period x Group interaction, F(2, 61) = 3.44, p
= .038, η2 = .10. Mean cortisol levels were significantly higher during the
examination period than during the no examination period (p < .001). During
both sessions, salivary cortisol significantly decreased after the intervention
(baseline vs. after intervention: p < .001) and returned to its initial level after the
SECPT (after intervention vs. after SECPT: p < .001). There were no signifi-
cant differences in cortisol levels between the three groups during the no exami-
nation period (all ps > .05). During the examination period, participants in the
nature group had significantly lower cortisol levels compared with the video
group after the intervention (p = .046), and this difference remained significant
after controlling for baseline cortisol levels (p = .044). No other group differ-
ences were found (all ps > .05). Mean cortisol levels are shown in Figure 4.
HR and HRV. In the analyses of HR as a dependent variable, there was a sig-
nificant main effect for Time, F(3, 141) = 57.16, p < .001, η2 = .55, a
Figure 4. Mean cortisol levels during the no examination and examination period,
in the nature-, gym- and video-group, respectively.
Note. Data are expressed as mean ± Standard Error of the Mean (SEM).
*p < .05 nature versus video.
Olafsdottir et al. 15
significant Time x Group interaction, F(6, 141) = 9.24, p < .001, η2 = .28,
and a significant interaction between Period and Time, F(3, 141) = 3.94, p =
.013, η2 = .08. No other main effects or interactions emerged (all Fs < 1.65,
all ps > .15). Groups did not differ in baseline HR levels (all ps > .05). After
the intervention, mean HR in the nature group was significantly higher com-
pared with mean HR in the video group (no exams: p = .017; exams: p =
.011), even after controlling for baseline values (no exams: p < .001; exams:
p < .001). No significant differences were found between the nature group
and the gym group (no exams: p = .244; exams: p = .064) or between the
gym group and the video group (no exams: p = .135; exams: p = .378).
There were no significant group differences during the SECPT and at follow-
up (all ps > .05). Mean HR data are depicted in Figure 5.
In the analyses of HFnu, a significant Time x Group interaction, F(6,
141) = 2.73, p = .015, η2 = .10 occurred. No other main effects or interac-
tions were found (all Fs < 2.21, all ps > .09). There was no significant
difference in HFnu between the three groups at baseline and after interven-
tion (all ps > .05).
Figure 5. Mean heart rate (expressed as beats per minute) during no exam
and exam period, in the green exercise group, gym group, and video group,
Note. Data are expressed as mean ± Standard Error of the Mean (SEM). Measurement points:
t1 = baseline; t2 = after intervention; t3 = during SECPT; t4 = follow-up. SECPT = Socially
Evaluated Cold-Pressor Test.
*p < .05 nature versus video.
16 Environment and Behavior 00(0)
During the examination period, participants in the video group showed
significantly higher HFnu during the SECPT compared with the nature group
(p = .037) and the gym group (p = .025), even after controlling for baseline
HFnu values (video vs. nature: p = .007; video vs. gym: p = .001). No other
group differences were found during the examination period or during the no
examination period (all ps > .05). Mean HFnu data are shown in Figure 6.
When NA was entered as a dependent variable, results showed significant
main effects for Period, F(1, 63) = 14.91, p < .001, η2 = .19 and Time,
F(2, 126) = 36.78, p < .001, η2 = .37, and a significant Period x Time
interaction, F(2, 126) = 10.80, p < .001, η2 = .15. There were no main or
interaction effects for Group (all Fs < 2.28, all ps > .11). NA was higher
during the examination period compared with the no examination period at
baseline (p < .001) and after the intervention (p = .001), but not at follow-
up (p = .364). NA scores decreased significantly from baseline to after
intervention across groups (no exams and exams: p < .001). No significant
Figure 6. Mean HFnu during no examination and examination period in the
nature-, gym- and video-group, respectively.
Note. Data are expressed as mean ± Standard Error of the Mean (SEM). Measurement
points: t1 = baseline; t2 = after intervention; t3 = during SECPT; t4 = follow-up. HF = high
frequency; SECPT = Socially Evaluated Cold-Pressor Test.
*p < .05 video versus nature and gym.
Olafsdottir et al. 17
changes in NA were found from after intervention to follow-up (no exams:
p = .515; exams: p = .933).
The analyses of PA showed a significant main effect for Time, F(2, 126) =
5.55, p = .005, η2 = .08 and a significant Time x Group interaction, F(4, 63) =
7.75, p < .001, η2 = .20. There was no difference in PA between the three groups
at baseline (all ps > .05). After the intervention, participants from the nature
group reported significantly higher PA scores compared with the gym group (no
exams: p = .016; exams: p = .004) and the video group (no exams and exams: p
< .001). These differences remained significant after controlling for baseline val-
ues (all ps < .05). No significant group differences were found for PA at follow-
up (all ps > .05). Mean NA and PA scores are shown in Table 1.
To the best of our knowledge, this is the first study to apply an ecologically
valid research design (a) to test whether walking in nature promotes more
restorative effects on mood and psychophysiological stress responses under
chronic and acute stress conditions compared with exercise alone and (b) to
disentangle the effects of walking and nature exposure by including all three
conditions, that is, walking in nature, walking on a treadmill, and nature
viewing. We expected participants to be more stressed during the examina-
tion period, and to benefit more profoundly when psychologically more
stressed, as has been previously suggested (Ulrich, 1981, 1983). We tested
these questions by measuring changes in endocrine markers of stress (corti-
sol) and indices of cardiac autonomic nervous system function (HR and HRV
Table 1. Means and Standard Deviations for Positive and Negative Affect Scores
During No Examinations and Examination Period, in the Nature Group, Gym
Group, and Video Group, Respectively.
No exams Exams
intervention Follow-up Baseline
Nature 13.11 (2.32) 11.17a (2.43) 12.56 (2.31) 15.83 (3.94) 11.78a (3.30) 12.00 (2.03)
Gym 15.13 (5.35) 12.96a (3.06) 13.42 (3.22) 17.21 (5.73) 14.50a (4.15) 14.08 (4.39)
Video 14.61 (3.50) 11.39a (1.80) 13.43 (4.44) 18.57 (6.49) 13.78a (3.83) 14.13 (4.30)
Nature 31.10 (5.97) 36.80bc (9.38) 34.90 (8.60) 31.30 (5.89) 37.50bc (8.37) 34.75 (8.77)
Gym 30.08 (6.53) 29.96 (6.50) 32.92 (5.21) 29.88 (6.41) 29.88 (6.42) 32.50 (5.10)
Video 32.04 (5.47) 26.04 (7.57) 30.96 (6.53) 32.13 (5.49) 26.57 (7.57) 31.43 (5.98)
Note. The values in parentheses are SDs.
aSignificant at the p < .001 level in comparison with baseline negative affect (NA).
bSignificant at the p < .05 level in comparison with the gym group.
cSignificant at the p < .001 level in comparison with the video group.
18 Environment and Behavior 00(0)
[HFnu]) as well as psychological affect, when life was relatively relaxed and
again during the examination period.
A first finding of this study concerns higher baseline cortisol levels during
the examination period compared with the no examination period. This is in
line with previous study findings on examination stress (Ignacchiti, Sesti-
Costa, Marchi, Schedraoui-Silva, & Mantovani, 2011) and suggests higher
stress levels during the examination time. Nevertheless, the order of no-exam-
ination and examination periods in the current study was fixed, so this effect
may have also been affected by seasonal variations, which primarily show
winter to summer decreases in cortisol levels (Hadlow, Brown, Wardrop, &
Henley, 2014; Hansen, Garde, Skovgaard, & Christensen, 2001; King et al.,
2000), probably associated with increased daylight exposure. As we found
higher cortisol levels even under conditions of increased daylight hours (from
4.25 hr in January to 17.73 hr in May; Saemundsson, 2013), it seems likely
that the observed increase in cortisol levels was due to examination stress,
which may have overridden any seasonal effects of reduced cortisol.
All interventions had the power to significantly decrease cortisol levels dur-
ing both life-stress periods, and there was no difference between the stress-
buffering effects on either chronic or acute stress when life was relatively
relaxed. Yet, when the individuals were under more stress, walking in nature
(nature group) resulted in the largest decrease in cortisol levels, and passive
viewing of nature scenes (video group) the least. This difference remained sig-
nificant even after controlling for initial baseline differences, indicating that
walking in nature may reduce stress levels to a greater extent when experienc-
ing higher real-life stress levels than just viewing nature. Nevertheless, there
was no statistically significant difference in average cortisol levels between
participants walking in nature or walking in the gym. There was no significant
difference in mean HR between the groups immediately after the interventions
although there was a trend for higher HR in the nature group compared with the
gym group despite being given the same instructions concerning exercise
intensity. The nature walk may, therefore, have been more challenging than the
gym walk in terms of physical effort. Moreover, after the intervention, partici-
pants in the nature group had significantly higher HR levels than the video
group (both periods), and a tendency for lower parasympathetic activity (as
indicated by the HF spectral component of HRV) than the other groups although
these differences did not reach statistical significance. HR reflects both effects
of sympathetic and parasympathetic activity. Given that the HF spectral com-
ponent was reduced, it is likely that the increase in HR in the nature group was
primarily due to decreased parasympathetic activity together with a simultane-
ous increase in sympathetic activity (Dong, 2016; Karemaker, 1999). HR was
lowest after the intervention in those who passively viewed nature.
Olafsdottir et al. 19
Previously, nature viewing has been shown to result in significant reduc-
tions in HR (Gladwell et al., 2012; Laumann et al., 2003; Ulrich et al., 1991)
and higher HFnu (Gladwell et al., 2012) and higher RMSSD as indicators of
parasympathetic activity after exposure to acute stress (Brown et al., 2013),
indicating faster recovery. The cold pressor test (without the social compo-
nent) increases HR in healthy participants (Jarvis, Okada, Levine, & Fu,
2015; Mourot, Bouhaddi, & Regnard, 2009; Sinha & Dubey, 2016) and
decreases HFnu (Aimie-Salleh & Malarvili, 2011). In the current study, this
was the case in both trial periods except for the video group. HFnu was higher
in the video group during the SECPT but significant only in the chronic stress
condition (exam period), suggesting that passive exposure to nature may help
individuals to cope with acute stress under such circumstances. There were
no differences between intervention groups in HR or HRV during the period
following the laboratory stressor (SECPT), which contrasts with the findings
by Brown et al. (2013) who found increased RMSSD during recovery from
acute mental stress after passive viewing of nature scenes. Yet, the stressors
used were different. Brown et al. (2013) used a task inducing mental load
(i.e., mental arithmetic), whereas the cold pressor task used in the current
study involves both physical and mental stress.
In summarizing the results on physiological responses, walking (regard-
less of environment/per se) seems to be most effective in buffering HPA axis
activation when under high and low chronic (life stress) but not acute stress.
Passive viewing of nature scenes, however, seems more effective in increas-
ing cardiac-vagal activation when faced with an acute stressor.
The current findings on cortisol are comparable with some extent to what
Park and colleagues (2010) found in their 24 experiments of 15 to 20 min
forest walks and views in Japan, although there are differences in research
design. As was the case with their results obtained from studies in forest envi-
ronments, the current findings on cortisol were not reflected in the changes
that occurred in cardiac-autonomic responses. Our protocol did not allow for
any resting period after the walks or viewings before the ECG was recorded,
as has been the case in the studies that demonstrated decreased sympathetic
and increased parasympathetic activation after active nature exposure (Li
et al., 2011; Park et al., 2010). This may be the reason why exercise with or
without nature exposure did not lead to such outcomes in the present study.
Mean HR after the intervention was higher in the nature group compared
with the gym group, although this difference was not statistically signifi-
cant. In spite of being given the same instructions concerning exercise
intensity, this may be an indication that the nature walk may have been
physically more challenging than the gym walk. Previous studies have
shown that self-selection of speed leads to slower walking inside than
20 Environment and Behavior 00(0)
outside (Ceri & Hassmen, 1991), and when exposed to nature, perception
of effort is reduced and mood enhanced (Akers et al., 2012).
All interventions resulted in significant reductions in NA for both periods.
Participants across all groups reported significantly higher NA during the
examination period, but only the nature group reported significant increases in
PA after the intervention in both periods. This finding indicates that participants
in the nature group felt better after the activity, which is in line with previous
study findings of the positive impact of nature exposure on mental health over
and above urban environments (Thompson-Coon et al., 2011). The current
findings add, however, to the current literature in that they show that individu-
als do not have to be under high life stress to benefit psychologically from
exposure to nature. According to Ulrich (1979, 1981), individual stress reduc-
tion occurs affectively and physiologically as the result of a perception of an
unthreatening environment. Physiological stress responses are mobilized by
perceived physical or psychological threats (the fight-flight response; Cannon,
1936) that put demands on the organism (Selye, 1955, 1979), which the brain
discerns as an emergency and seeks to deal with immediately. Yet, as demon-
strated by a large body of research on stress appraisal over the last five decades,
individuals have the capacity to respond very differently to a given stress situ-
ation, depending on their cognitive appraisal. Such appraisal processes can eas-
ily destabilize allostatic balance and sustain a stressful situation by engaging
with particular situations as an emergency, or worrying about things that might
or might not happen (Sapolski, 2004; Selye, 1979). As Ulrich (1981) argues,
the key to stress release is to refrain or distance oneself from negative thoughts
and appraise the situation as unthreatening. The current findings in terms of
increased PA and reduced NA and reduced cortisol seem to indicate that walk-
ing in nature is beneficial for this process, especially under conditions of
chronic or longer-lasting real-life stress.
There are several limitations to this study. First, participants were intro-
duced to the study’s settings and more thoroughly to the experimental proce-
dure immediately before the experiment. This might have been a cause for
stress as most of the study volunteers had never been physically present in these
settings, although they were aware of their existence and none had participated
in an experiment of this sort before. Different results may have been demon-
strated if we had followed Park and colleagues’ (2010) approach and intro-
duced the setting and procedure on the day before the experiment and put them
up in a guesthouse the night before with controlled food and beverage intake.
Second, the fixed order of non-examination and examination periods might
have affected some of the current findings. Nevertheless, and as discussed in
the previous paragraph, it is unlikely that seasonal effects would have con-
founded the findings on cortisol. Third, and also with regard to the cortisol data,
Olafsdottir et al. 21
10 min of pre-experimental rest may have been too short to obtain a meaningful
baseline. Nevertheless, the current study was primarily designed to assess cor-
tisol responses to an acute stressor. Future studies should also address the
impact of nature walking on other indices of HPA axis activity, such as diurnal
cortisol rhythmicity. Fourth, the weather in the nature setting varied and was at
times challenging although the forest provided some shelter from the wind and
rain. The conditions on the ground also varied. During the first period, the
ground was covered with snow and sometimes slippery in places. This may
have somewhat reduced participants’ enjoyment of the nature walk—although
they were warned—as they tried to tackle the inconvenience of a challenging
environment. However, it needs to be stated that no one fell or reported injuries.
On the contrary, the research diary revealed that most people enjoyed the walk
and returned from it feeling better as indicated by the changes in NA and PA.
Fifth, both the nature setting and the gym were open for the public while the
study was carried out, which meant that other people were present, which may
have affected the results. Sixth, we did not control for exercise intensity
between the nature and the gym walking groups (e.g., setting the treadmill at a
1% incline to represent outdoor walking, measuring exercise intensity) other
than (a) instructing participants to walk at a leisurely pace that they felt com-
fortable with and (b) setting the pace of the treadmill (gym group) to a level to
match the distance to be covered on the nature walk (approx. 4 km) in 40 min.
More controlled conditions would have been desirable in terms of method-
ological rigor, thereby improving internal validity; however, increased levels of
standardization usually reduce ecological validity and, therefore, impede the
generalizability of results. Nevertheless, future studies should assess perceived
exertion and measure exercise intensity as objectively as possible, for example,
by including accelerometer/pedometer data. Seventh, it should be stressed that
the HR and HRV findings should be regarded with caution due to the missing
ECG data and small sample size (N = 50). Finally, a 40-min intervention may
have been too short to induce more significant changes in stress responses. Yet,
it is longer than the 10 to 30 min interventions used in previous studies investi-
gating the effects of nature exposure on stress responses (see, for example,
Brown et al., 2013; Li et al., 2011; Park et al., 2010; Ulrich et al., 1991).
Furthermore, studies on the health effects of nature exposure have shown that
even short visits to nature (30 min) can decrease cortisol levels and perceived
stress (Tyrväinen et al., 2014), and the biggest impact on mental well-being
seems to occur during the first 5 min (Barton & Pretty, 2010).
Although there are exceptions (Brown et al., 2013), previous studies inves-
tigating the effects of nature exposure on stress responses tended to monitor
responses to a single laboratory stressor, and then monitor recovery after
exposure to nature (Ulrich et al., 1991). The current study, however, extends
22 Environment and Behavior 00(0)
this approach by assessing the effects of nature exposure on both chronic,
general life stress and acute stress. The results suggest that periodic passive or
active nature exposure or exercise alone (walking) might generally be benefi-
cial in reducing stress levels. Although our hypotheses can only be partially
confirmed, we found a tendency in favor of nature exposure in all parameters
investigated. The results indicate that—when under high periodic life stress—
stress responses can be mitigated more effectively by a 40-min walk in nature,
than by physical exercise alone or by watching nature scenes (in this order).
When life stress was more profound, the video group showed significantly
higher HFnu reflecting cardiac-vagal activation during acute stress, which
suggests that relaxed, passive nature viewing may help people deal with acute
stress more readily than physical exercise with or without nature exposure.
Overall, the findings have important implications for local authorities to
advocate the therapeutic agency of nature walks and views and to provide the
public with easy access to nature-rich places where people can relax either
passively by viewing nature or actively by walking in nature.
We thank the Department of Physiology, University of Iceland, for housing the
fieldwork in Iceland. Much gratitude is also extended to Sveinbjörn Yngvi Gestsson,
Sascha Helsper, Logi Jónsson, Lilja Guðrún Steinsdóttir, Anna Guðmundsdóttir,
Ágústa Johnson, Hreyfing ehf, Helgi Gíslason, Reykjavik Forestry Association, Jón
Sigurdsson, Nicole Knoblauch, Elizabeth Cook, and the National University
Hospital of Iceland for supporting this research. Unnur Valdimarsdottir, University
of Iceland, and Marcia Worrell, Roehampton University, had input at the develop-
ment stage of the study.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research,
authorship, and/or publication of this article.
The author(s) disclosed receipt of the following financial support for the research,
authorship, and/or publication of this article: This study was funded by the Fonds
National de la Recherche Luxembourg (AFR ref. 3965162) and the University of
Luxembourg (Internal Research Funding F3R-INS-PUL-13BREA/BREATH).
Gunnthora Olafsdottir https://orcid.org/0000-0002-4250-6822
André Schulz https://orcid.org/0000-0002-9381-2651
Olafsdottir et al. 23
Aimie-Salleh, N., & Malarvili, M. B. (2011). Study of relationship between heart rate
variability and autonomic function using cold pressor test for Malaysian popula-
tion. IEEE Colloquium on Humanities, Science and Engineering Research, 5-6,
Akers, A., Barton, J., Cossey, R., Gainsford, P., Griffin, M., & Micklewright, D.
(2012). Visual color perception in green exercise: Positive effects on mood
and perceived exertion. Environmental Science & Technology, 46, 8661-8666.
Barton, J., & Pretty, J. (2010). What is the best dose of nature and green exercise
for improving mental health? A multi-study analysis. Environmental Science &
Technology, 44, 3947-3955. doi:10.1021/es903183r
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 envi-
ronments. BMC Public Health, 10, Article 456. doi:10.1186/1471-2458-10-456
Brown, D. K., Barton, J. L., & Gladwell, V. F. (2013). Viewing nature scenes posi-
tively affects recovery of autonomic function following acute-mental stress.
Environmental Science & Technology, 47, 5562-5569. doi:10.1021/es305019p
Cannon, W. (1936, May). The role of emotion in disease. Annals of Internal
Medicine, 9(11), 1453-1465.
Ceri, R., & Hassmen, P. (1991). Self-monitored exercise at three different RPE inten-
sities in treadmill vs. field running. Medical Science in Sport and Exercise, 23,
Chida, Y., & Steptoe, A. (2009). Cortisol awakening response and psychosocial fac-
tors: A systematic review and meta-analysis. Biological Psychology, 80, 265-
Crawford, J. R., & Henry, J. D. (2004). The positive and negative affect schedule
(PANAS): Construct validity, measurement properties and normative data in a
large non-clinical sample. British Journal of Clinical Psychology, 43, 245-265.
Dong, J. G. (2016). The role of heart rate variability in sports physiology (review).
Experimental and Therapeutic Medicine, 11, 1531-1536. doi:10.3892/
Driver, B. L. (1976). Quantification of outdoor recreationists’ preferences. In B.
Smissen & J. Myers (Eds.), Research: Camping and environmental educa-
tion (HPEP Series No. 11, pp. 165-187). University Park: Pennsylvania State
Gladwell, V. F., Brown, D. K., Barton, J. L., Tarvainen, M. P., Kuoppa, P., Pretty,
J., . . . Sandercock, G. R. H. (2012). The effects of views of nature on autonomic
control. European Journal of Applied Psychology, 112, 3379-3386. doi:10.1007/
Gladwell, V. F., Brown, D. K., Wood, C., Sandercock, G. R., & Barton, J. L. (2013).
The great outdoors: How a green exercise environment can benefit all. Extreme
Physiology & Medicine, 2, Article 3. doi:10.1186/2046-7648-2-3
24 Environment and Behavior 00(0)
Hadlow, N. C., Brown, S., Wardrop, R., & Henley, D. (2014). The effects of
season, daylight saving and time of sunrise on serum cortisol in a large popu-
lation. Chronobiology International, 31, 243-251. doi:10.3109/07420528.20
Hansen, Å. M., Garde, A. H., Skovgaard, L. T., & Christensen, J. M. (2001). Seasonal
and biological variation in urinary epinephrine, norepinephrine, and corti-
sol in healthy women. Clinica Chimica Acta, 309, 25-35. doi:10.1016/S0009-
Hartig, T., Mitchell, R., de Vries, S., & Frumkin, H. (2014). Nature and health.
Annual Review of Public Health, 35, 207-228. doi:10.1146/annurev-publ-
Health Council of the Netherlands. (2004). Nature and health: The influence of nature
on social, psychological and physical well-being. The Hague: Health Council
of the Netherlands/Dutch Advisory Council for Research on Spatial Planning,
Nature and the Environment. Retrieved from http://www.gezondheidsraad.nl/
Herzog, T. R., Black, A. M., Fountaine, K. A., & Knotts, D. J. (1997). Reflection and
attentional recovery as distinctive benefits of restorative environments. Journal
of Environmental Psychology, 17, 165-170. doi:10.1006/jevp.1997.0051
Home, R., Hunziker, M., & Bauer, N. (2012). Psychosocial outcomes as motivations
for visiting nearby urban green spaces. Leisure Science, 34, 350-365. doi:10.108
Ignacchiti, M. D. C., Sesti-Costa, R., Marchi, L. F., Schedraoui-Silva, S., &
Mantovani, B. (2011). Effect of academic psychological stress in post-graduate
students: The modulatory role of cortisol on superoxide release by neutrophils.
Stress, 14, 290-300. doi:10.3109/10253890.2010.545459
James, P., Banay, R. F., Hart, J. E., & Laden, F. (2015). A review of the health ben-
efits of greenness. Current Epidemiology Reports, 2, 131-142. doi:10.1007%
Jarvis, S. S., Okada, Y., Levine, B. D., & Fu, Q. (2015). Central integration and
neural control of blood pressure during the cold pressor test: A comparison
between hydrochlorothiazide and aliskiren. Physiological Reports, 3(9), e12502.
Kaplan, R., & Kaplan, S. (1989). The experience of nature: A psychological perspec-
tive. New York, NY: Cambridge University Press.
Kaplan, S. (2001). The restorative benefits of nature: Toward an integrative frame-
work. Journal of Environmental Psychology, 15, 169-182. doi:10.1016/0272-
Kaplan, S., & Talbot, J. F. (1983). Psychological benefits of a wilderness experience.
In I. Altman & J. F. Wohlwill (Eds.), Human behavior and the environment:
Advances in theory and research 6 (pp. 163-204). New York, NY: Plenum Press.
Karemaker, J. M. (1999). Autonomic integration: The physiological basis of cardio-
vascular variability. Journal of Physiology, 517(Pt. 2), Article 316. doi:10.1111/
Olafsdottir et al. 25
King, J. A., Rosal, M. C., Ma, Y., Reed, G., Kelly, T.-A., Stanek, E. J., & Ockene, I. S.
(2000). Sequence and seasonal effects of salivary cortisol. Behavioral Medicine,
26(2), 67-73. doi:10.1080/08964280009595753
Knopf, R. C. (1987). Human behavior, cognition and affect in the natural environ-
ment. In D. Stokos & I. Altman (Eds.), Handbook of environmental psychology
(pp. 783-825). New York, NY: John Wiley.
Kondo, M., Fluehr, J., McKeon, T., & Branas, C. (2018). Urban green space and its
impact on human health. International Journal of Environmental Research and
Public Health, 15(3), Article 445. doi:10.3390/ijerph15030445
Kudielka, B. M., & Wüst, S. (2010). Human models in acute and chronic stress:
Assessing determinants of individual hypothalamus-pituitary-adrenal axis activ-
ity and reactivity. Stress, 13, 1-14. doi:10.3109/10253890902874913
Lass-Hennemann, J., Deuter, C. E., Kuehl, L. K., Schulz, A., Blumenthal, T. D., &
Schachinger, H. (2010). Effects of stress on human mating preferences: Stressed
individuals prefer dissimilar mates. Proceedings of the Royal Society, Series B:
Biological Sciences, 277, 2175-2183. doi:10.1098/rspb.2010.0258
Laumann, K., Garling, T., & Stormark, K. (2003). Selective attention and heart
rate responses to natural and urban environments. Journal of Environmental
Psychology, 23, 125-134. doi:10.1016/S0272-4944(02)00110-X
Li, Q., Otsuka, T., Kobayashi, M., Wakayama, Y., Inagaki, H., & Katsumata, M.
(2011). Acute effects of walking in forest environments on cardiovascular and
metabolic parameters. European Journal of Applied Physiology, 111, 2845-2853.
Mathew, B. C., Biju, R. S., & Thapalia, N. (2005). An overview of electrochemilu-
minescent (ELC) technology in laboratory investigations. Kathmandu University
Medical Journal, 3(1), 91-93.
Mourot, L., Bouhaddi, M., & Regnard, J. (2009). Effects of the cold pressor test on
cardiac autonomic control in normal subjects. Physiological Research, 58, 83-91.
Ohly, H., White, M. P., Wheeler, B. W., Bethel, A., Ukoumunne, O. C., Nikolaou,
V., & Garside, R. (2016). Attention restoration theory: A systematic review of
the attention restoration potential of exposure to natural environments. Journal
of Toxicology and Environmental Health, Part B, 19, 305-343. doi:10.1080/109
Olafsdottir, G., Cloke, P., & Vögele, C. (2017). Place, green exercise and stress: An
exploration of lived experience and restorative effects. Health & Place, 46, 358-
Park, B., Tsunetsugu, Y., Kasarani, T., Kagawa, T., & Miyazaki, Y. (2010). The phys-
iological effects of Shinrin-yoku (taking in the forest atmosphere or forest bath-
ing): Evidence from field experiments in 24 forests across Japan. Environmental
Health and Preventive Medicine, 15, 18-26. doi:10.1007/s12199-009-0086-9
Parsons, R. (1991). Recovery from stress during exposure to videotaped outdoor envi-
ronments (Doctoral dissertation). University of Arizona, Tucson.
Pretty, J., Griffin, M., Sellens, M., & Pretty, C. (2003). Green exercise: Complementary
roles of nature, exercise and diet in physical and emotional well-being and
26 Environment and Behavior 00(0)
implications for public health policy [CES occasional paper]. Colchester, UK:
University of Essex.
Pretty, J., Peacock, J., Sellens, M., & Griffin, M. (2006). The mental and physi-
cal health outcomes of green exercise. International Journal of Environmental
Health Research, 15, 319-337. doi:10.1080/09603120500155963
Rossi, V., & Pourtois, G. (2012). Transient state-dependent fluctuations in anxiety
measured using STAI, POMS, PANAS or VAS: A comparative review. Anxiety
Stress Coping, 25, 603-645. doi:10.1080/10615806.2011.582948
Saemundsson, T. (Ed.). (2013). Almanak fyrir Ísland [The calendar for Iceland], 177.
Reykjavik: University of Iceland.
Sapolski, R. M. (2004). Why zebras don’t get ulcers: The acclaimed guide to stress,
stress-related diseases, and coping (3rd ed.). New York, NY: St. Martin’s Griffin.
Schwabe, L., Haddad, L., & Schächinger, H. (2008). HPA axis activation by a socially
evaluated cold-pressor test. Psychoneuroendocrinology, 33, 890-895.
Selye, H. (1955). Stress and disease. Science, 122, 625-631. doi:10.1126/sci-
Selye, H. (1979). The stress of my life. New York, NY: Van Nostrand.
Sinha, B., & Dubey, D. K. (2016). Blood pressure variability and baroreflex sensitivity
of a healthy male during cold pressor test that induced development of neurocar-
diogenic syncope. Journal of Basic and Clinical Physiology and Pharmacology,
27, 437-443. doi:10.1515/jbcpp-2015-0004
Steptoe, A., & Vögele, C. (1991). The methodology of mental stress testing in cardio-
vascular research. Circulation, 83(Suppl. 4), II14-II24.
Task Force of the European Society of Cardiology and the North American Society of
Pacing Electrophysiology. (1996). Heart rate variability: Standards of measure-
ment, physiological interpretation and clinical use. Circulation, 93, 1043-1065.
Thompson-Coon, J., Boddy, K., Stein, K., Whear, R., Barton, J., & Depledge, M. H.
(2011). Does participating in physical activity in outdoor natural environments
have a greater effect on physical and mental wellbeing than physical activity
indoors? A systematic review. Environmental Science & Technology, 45, 1761-
Tyrväinen, L., Ojala, A., Korpela, K., Lanki, T., Tsunetsugu, Y., & Kagawa, T.
(2014). The influence of urban green environments on stress relief measures: A
field experiment. Journal of Environmental Psychology, 38, 1-9. doi:10.1016/j.
Ulrich, R. S. (1979). Visual landscapes and psychological well-being. Landscape
Research, 4(1), 17-23. doi:10.1080/01426397908705892
Ulrich, R. S. (1981). Natural versus urban scenes: Some psycho-physiological effects.
Environment & Behavior, 13, 523-556. doi:10.1177/0013916581135001
Ulrich, R. S. (1983). Aesthetic and affective response to natural environment. In I.
Altman & J. F. Wohlwill (Eds.), Behavior and the natural environment (pp. 85-
125). New York, NY: Plenum Press.
Ulrich, R. S. (1984). View through a window may influence recovery from surgery.
Science, 224, 420-421. doi:10.1126/science.6143402
Olafsdottir et al. 27
Ulrich, R. S. (1986). Human responses to vegetation and landscapes. Landscape and
Urban Planning, 13, 29-44. doi:10.1016/0169-2046(86)90005-8
Ulrich, R. S., Simons, R. F., 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. doi:10.1016/S0272-
Valtchanov, D., Barton, K. R., & Ellard, C. (2010). Restorative effects of virtual
nature settings. Cyberpsychology, Behavior, and Social Networking, 13, 503-
Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of
brief measures of positive and negative affect: The PANAS Scales. Journal of
Personality and Social Psychology, 47, 1063-1070.
Gunnthora Olafsdottir, PhD, is the director of research at the Icelandic Tourist
Board and a visitor research fellow at University of Exeter and University of
Luxembourg. Her research interests include restorative environments, human-nature
relations, and human health and well-being in the context of leisure and tourism.
Paul Cloke, PhD, is a professor of human geography at University of Exeter and the
founding editor of the international and multidisciplinary academic Journal of Rural
Studies, published by Elsevier Science. He has research interests in social and cultural
geographies of rurality, nature-society relations, ethics and care, and landscapes of
André Schulz, PhD, is a research scientist at University of Luxembourg with research
interests in the psychoneuroendocrinology of stress and interoception.
Zoé van Dyck, PhD, is a research associate at University of Luxembourg. Her
research interests include cardiovascular and respiratory systems, human behavior,
treatment, and clinical psychology.
Thor Eysteinsson, PhD, is director of Department of Physiology and professor of
neurophysiology at the Faculty of Medicine, University of Iceland. His work focuses
primarily on electrophysiology and eye research.
Björg Thorleifsdottir is an assistant professor of physiology at University of Iceland
with primary research interests in biorhythms and sleep.
Claus Vögele, PhD, is a professor of health and clinical psychology at the University
of Luxembourg. His main research areas include biological psychology, psychophysi-
ology, and behavioral medicine where he examines psychocardiology, human behav-
ior/decision-making and its neurobiological correlates, emotion regulation, interocep-
tion, and stress.