Content uploaded by Ruth Kjærsti Raanaas
Author content
All content in this area was uploaded by Ruth Kjærsti Raanaas on Mar 18, 2016
Content may be subject to copyright.
Article
Health benefits of a view of nature
through the window: a quasi-
experimental study of patients in a
residential rehabilitation center
Ruth Kjærsti Raanaas
1
, Grete Grindal Patil
1
and Terry Hartig
2
Abstract
Objective: To examine the health benefits of a bedroom window view to natural surroundings for patients
undergoing a residential rehabilitation programme.
Design: Longitudinal quasi-experiment.
Setting: A residential rehabilitation centre.
Subjects: Two-hundred and seventy-eight coronary and pulmonary patients provided data at all measure-
ment points during the programme.
Intervention: Blind, quasi-random allocation to a private bedroom with a panoramic view to natural
surroundings or with a view either partially or entirely blocked by buildings.
Main measures: Self-reported physical and mental health (SF-12), subjective well-being, emotional states,
use of the private bedroom and leisure activities.
Results: For women, a blocked view appeared to negatively influence change in physical health
(time 3view 3gender interaction, F(4,504) ¼2.51, P¼0.04), whereas for men, a blocked view appeared
to negatively influence change in mental health (time 3view 3gender interaction, F(4,504) ¼5.67, P<0.01).
Pulmonary patients with a panoramic view showed greater improvement in mental health than coronary
patients with such a view (time 3view 3diagnostic group interaction, F(4,504) ¼2.76, P¼0.03). Those with
a panoramic view to nature more often chose to stay in their bedroom when they wanted to be alone than
those with a blocked view (odds ratio (OR) ¼2.32, 95% confidence interval (CI) 1.08–5.01).
Conclusion: An unobstructed bedroom view to natural surroundings appears to have better supported
improvement in self-reported physical and mental health during a residential rehabilitation programme,
although the degree of change varied with gender and diagnostic group.
Keywords
Cardiac rehabilitation, pulmonary rehabilitation, leisure, recreation, psychological factors
Received: 9 May 2011; accepted: 14 May 2011
1
Department of Plant and Environmental Sciences, Norwegian
University of Life Sciences, A
˚s, Norway
2
Institute for Housing and Urban Research, Uppsala University,
Ga
¨vle, Sweden
Corresponding author:
Ruth Kjærsti Raanaas, Department of Plant and Environmental
Sciences, Norwegian University of Life Sciences, Box 5003, N-
1432 A
˚s, Norway
Email: ruth.raanaas@umb.no
Clinical Rehabilitation
26(1) 21–32
ÓThe Author(s) 2011
Reprints and permissions:
sagepub.co.uk/journalsPermissions.nav
DOI: 10.1177/0269215511412800
cre.sagepub.com
Introduction
Features of the physical environment can
promote desirable patient outcomes in health
care institutions.
1–4
One such feature is the con-
nection between the indoor and outdoor environ-
ment provided by a window in a patient’s
bedroom. Fresh air and sunlight have long been
thought to promote patient well-being,
5
but until
relatively recently there was little experimental
evidence that what patients could see from their
room also could be beneficial. In a pioneering
study, however, Ulrich
6
compared recovery
from surgery among patients in hospital rooms
with a view of trees versus a brick wall. Those
with the tree view used less potent analgesics,
received fewer negative comments in nurses’
notes and had shorter postoperative hospital
stays.
To explain these benefits, Ulrich
6
proposed
that a view of natural elements serves as a dis-
traction that evokes positive emotions, counter-
acting stress and enhancing pain management.
His study has not been replicated, but several
experiments have addressed components of his
explanation.
7–12
Studies in clinical settings have
reported beneficial effects of indoor interventions
involving nature imagery.
8–10
For example,
Diette et al.
8
reported better pain control
among patients undergoing flexible bronchos-
copy while lying on a bed surrounded by a cur-
tain with a large landscape scene, in comparison
to patients without such a ‘bedscape’.
Experiments with non-clinical samples have
found enhanced psychophysiological stress
reduction with a window view to nature versus
no view
11
and videotapes of natural settings
versus urban and pedestrian mall settings.
12
The present study extends this line of research
by assessing the effects of window view contents
on patients in a residential rehabilitation centre.
The focus on residential centres is of interest
because many of them are located in natural
surroundings and offer window views onto those
surroundings.
5,13,14
Even with seemingly easy
physical access to the natural surroundings,
window views may have considerable importance,
particularly if harsh weather conditions and poor
health keep patients indoors.
In general, patients undergoing rehabilitation
are a vulnerable group, with compromised health
presenting a major life crisis.
15,16
Programmes
typically last for multiple weeks and are intended
to help patients restore physical function and
learn about lifestyle change and how to live
with their condition. Participants in a rehabilita-
tion programme have a number of mandatory
activities during the day that may be physically
and mentally demanding. Thus, although the
rehabilitation programme is meant to be restor-
ative in a sense, patients may frequently experi-
ence it as stressful. Withdrawal to a private room
with a view to natural surroundings can support
stress recovery on such occasions. We therefore
hypothesize that availability of a private room
with a window view to natural surroundings
will cumulatively have positive effects on rehabil-
itation patients’ health and well-being.
Given this hypothesis, it is important to mea-
sure how patients actually use their bedrooms,
and in particular the extent to which they use it
when they want to withdraw and be alone versus
using it only for sleeping. We hypothesize that
patients with a window view towards the natural
surroundings will more frequently use their room
as a place for contemplation and waking rest.
There are many possibilities for voluntary leisure
time activities at a rehabilitation centre, such as
social activities, watching TV and going out for
walks. We measured the performance of these
activities as well, to see for example whether
patients in rooms without a window view to
nature would spend more of their spare time
watching television whereas those with such a
view would stay more in their rooms. We also
assessed satisfaction with the window view and
the design of the private room.
Methods
We conducted a longitudinal quasi-experiment
at the rehabilitation centre in Røros, Norway.
The main experimental factor comprised three
types of window view from patient rooms,
22 Clinical Rehabilitation 26(1)
which varied in the openness of the view to the
natural surroundings. The study was coordi-
nated with a separate study on the effect of
indoor foliage plants on patient health and
well-being.
17
In that study we compared the
self-reported health and well-being of patients
during 11 months before and 11 months after
many large plants were placed in common
areas around the centre in November, 2008.
The present study examined the effects of the
window view from the private bedrooms of the
same patients, where there were no plants either
before or after the plant intervention in the
common areas.
The Røros Centre offers rehabilitation pro-
grammes for different diagnostic groups. These
include patients with a coronary diagnosis (e.g.
have undergone heart surgery or have a history of
coronary heart disease) and patients with a pulmo-
nary diagnosis (e.g. asthmatic diseases or chronic
obstructive pulmonary disease). Hereinafter we
will refer simply to ‘heart’ and ‘lung’ patients, in
line with the terms used by centre staff.
Patients are referred to the centre by their
regular physician and have been at home prior
to the stay; they typically do not enter directly
from acute care facilities. Rehabilitation pro-
grammes are run with groups of at most 20
heart or lung patients, and the length of the
programme is normally four weeks. The groups
are cared for by teams of staff members from
different professions (physical therapy, occupa-
tional therapy, nutrition, bioengineering, nurs-
ing, medicine).
During 2007–2008, patients preparing for a
stay at the centre were invited to participate in
the study via a letter sent with an initial question-
naire by a secretary at the centre. Potential par-
ticipants were informed that it was a study about
the aesthetics of the centre in relation to health
and well-being among patients there. They were
also informed that participation was voluntary.
No further information was given about the envi-
ronmental factors being measured. We obtained
informed consent from those patients willing to
participate. The study was approved by the
Regional Committee for Medical and Health
Research Ethics and the privacy ombudsman
for research at the Norwegian Social Science
Data Service.
Room assignments were administered by a
secretary at the centre (not the same one who
mailed the invitation to participate in the
study). Following standard procedure at the
centre, allocation was stratified by diagnostic
group, as different sets of rooms were reserved
for the different groups. For each diagnostic
group, incoming patients were allocated to the
rooms according to the alphabetical order of
their last names, starting with the lowest room
number. The few rooms with a totally blocked
view were among the lowest room numbers in
both groups. The secretary who made the alloca-
tion to rooms, as well as the other staff members
who interacted with the patients, were not aware
of the window view factor being measured.
Study site and description of window views
The rehabilitation centre was built in the moun-
tain village of Røros in 2003 for the Norwegian
National Association for Heart and Lung
Disease. The centre has modern architecture,
with large open spaces, large windows, light col-
ours, and extensive use of wood and stone in the
interior. Window views from the common areas
are dominated by grass and low trees, a parking
area, and other parts of the building or neigh-
bouring buildings. In contrast, window views
from the bedrooms encompass the valley and
the mountains, though in some rooms the views
of the landscape beyond the centre are partially
or entirely blocked by sections of the building
itself or by other buildings.
Three view categories were created on the
basis of the best possible (least obstructed) view-
point from inside the room. In the panoramic
view condition, rooms had an unobstructed
view onto the valley and mountains, with no
buildings blocking the view. In the partially
blocked view condition, rooms had a view to
the landscape that was partly blocked by parts
of the centre and/or other buildings. In the
blocked view condition, the view to the landscape
Raanaas et al. 23
was completely blocked by other parts of the
centre. Natural elements such as grass could be
seen outside, but there was no view to the land-
scape surrounding the centre.
Rooms at the centre are furnished with a bed,
sofa, desk, and small table. A chair near the
window can easily be oriented toward the view
the patient considers best. The room interior is
identical for all rooms, although the floor plan of
some of the rooms mirrors that of others. The
heart and lung patients used 52 rooms alto-
gether. On a few occasions, rooms primarily
used by heart patients were used by lung patients
and vice versa, so that the heart patients used a
total of 27 different rooms and the lung patients
used a total of 29 different rooms. Five of the
rooms are somewhat larger than the others and
these were used by patients in both diagnostic
groups. Two of them have panoramic views
and three have partially blocked views.
Measurement procedures and
outcome measures
The participants received questionnaires at five
different time points: with the initial invitation
letter two weeks before arrival at the centre
(T0), at the time of arrival (T1), two weeks
after arrival (T2), at the end of the four-week
programme (T3), and two weeks after returning
home (T4). The secretary who mailed out the
invitation letter with the T0 questionnaires also
maintained the coding key with participants’
names and registration codes, ensuring their ano-
nymity. She was also responsible for mailing out
the last questionnaire (T4). A member of the care
team working with the given group of patients
handed out the questionnaires for T1, T2 and T3
with instructions to fill them out and return them
in a mailbox placed in the centre’s reception area.
The staff were instructed not to inform the par-
ticipants about the interest in specific environ-
mental factors at the centre (which the staff
took to mean the indoor plants). The completed
forms were forwarded to the research team, and
the data were entered by a research assistant
blind to the window view conditions of the
participants. Information about the room
number for all participants who filled in at least
one questionnaire was collected from the secre-
tary holding the coding key after data collection
concluded. Patient assignments to view catego-
ries were derived from knowledge about their
room number. Room numbers were linked with
view categories on the basis of observations
made by the research team on-site.
Measures of health and well-being were
included in the questionnaires administered at
T0 through T4. The 12-item Short Form
Health Survey (SF-12) measures self-perceived
physical and mental health.
18
Responses were
scored according to the instructions provided in
the SF-12 version 2 manual.
19
Scores can range
from 0 to 100, with higher scores indicating
better health. Cronbach’s alpha ranged from
0.86 to 0.88 across the measurement points.
Subjective well-being was measured with a
single item: ‘When you look at your life in
general, how happy would you say that you
are?’
20
Responses were given with an 11-point
scale (0 ¼extremely unhappy, 10 ¼extremely
happy).
Current emotional well-being was measured
using items from a measure of affect based on
the circumplex model.
21
We focused on acti-
vated–unpleasant (AUP) and unactivated–
pleasant (UAP) affects. AUP affect was
measured with the adjectives annoyed, fearful,
jittery and anxious. UAP was measured with
the terms relaxed, content, at rest, and calm.
Instructions referred to the degree to which
these emotions were experienced at the moment
(1 ¼not at all; 5 ¼very much). The two measures
AUP and UAP were combined after reversal of
scores on the AUP measure, since analyses
of each separately showed that their results mir-
rored each other. Cronbach’s alpha ranged
from 0.83 to 0.87 across the measurement points.
The T3 questionnaire also included specific
questions about the extent to which the partici-
pant was satisfied with the window view and the
design of the private bedroom. Responses were
given on 5-point scales (0 ¼not at all; 4 ¼very
much).
24 Clinical Rehabilitation 26(1)
The T4 questionnaire included questions
about how many hours the participant spent in
the bedroom during the day and for what
purposes they used the bedroom. Responses to
the latter question were recoded into two catego-
ries: only for sleeping during the night and for
taking a nap during the day (coded 0), and for
withdrawing and being alone as well (coded 1).
The T4 questionnaire also included a question
about how often the participant went out
for walks during the four-week period (0 ¼never,
1¼once or twice during the stay,2 ¼once or twice
every week, 3 ¼every second day, 4 ¼every day).
Responses were recodedinto two categories: up to
twice during the stay at the centre (coded 0), and
more than twice during the stay (coded 1).
The T4 questionnaire further included items
about whether or not the participant spent time
in the reception area, TV rooms, and corridor
resting areas during daytime, evenings and week-
ends (0 ¼no, 1 ¼yes for each of the nine ques-
tions for the three areas 3three periods). For
each common area, a mean was calculated for
each participant looking across the three differ-
ent periods. The mean scores could range from 0
to 1, where 1 indicates more use of that particular
common area. The common area measures
reflect on the degree of social activity, but the
use of the TV rooms is assumed to reflect on
the amount of TV watching.
A variety of socio-demographic variables
were measured at T0. Those relevant for the pre-
sent study were gender and age in years.
Statistical analyses
Imputations for missing data were made at the
item level using the expectation maximization
method,
22
with data for other items from the
same questionnaire at the same time point used
for reference.
One-way ANOVA were used to assess the
effects of the window view on satisfaction with
the window view and interior room design.
Repeated-measures ANOVA were used to
assess the effects of the window view on change
in the self-reported health, subjective well-being,
and emotion outcomes from T1 to T3. For all of
these analyses, gender, diagnostic group and
plant intervention period (pre- and post-) were
treated as additional between-subjects factors.
Age and the corresponding measure from T0
were used as covariates. Greenhouse-Geisser
adjusted degrees of freedom and associated prob-
abilities are reported where appropriate for the
repeated measures ANOVA. The type III sum
of squares was used, given unequal cell numbers.
Logistic regression was used to assess the odds
that the patients performed some leisure activi-
ties more often if they had a particular window
view, independent of age, gender, diagnostic
group, and plant intervention.
Results
During the study period, 250 lung and 345 heart
patients (in 16 lung and 20 heart groups) com-
pleted at least three of the four weeks in one of
the centre’s rehabilitation programmes and were
eligible for the study. The flow of patients is
shown in Figure 1. Four hundred and eighty-
one patients consented to participate. Of these,
278 answered all three questionnaires at the
centre (T1–T3) and were included in analyses of
change in the health and other outcomes. The
questions posed at T4 about activities at the
centre were answered by 380 patients. Patients
did not complete all questionnaires because of
unwillingness (e.g. due to fatigue) and on a few
occasions because they did not receive the
questionnaire from centre staff.
The demographic characteristics of the groups
are presented in Table 1. The mean age of the
participants did not differ across the view condi-
tions (P¼0.27). Gender was not significantly
associated with view type (x
2
¼2.89, P¼0.24).
However, more of the heart than the lung
patients had panoramic views (x
2
¼10.37,
P<0.01). Relatively more of the participating
patients had a panoramic view during the
second year of the study, after initiation of the
plant intervention, whereas fewer had a partially
blocked view during the same period (x
2
¼7.20,
P<0.05). These differences further motivate the
Raanaas et al. 25
Table 1. Baseline demographic data for patients in the different window view conditions
Panoramic Partially blocked Blocked
Age in years, mean (SD) 62 (8.6) 64 (8.6) 62 (9.1)
Sex
Male N(%) 83 (73) 86 (64) 18 (60)
Female N(%) 31 (27) 48 (36) 12 (40)
Diagnostic group
Heart N(%) 81 (71) 69 (51) 16 (53)
Lung N(%) 33 (29) 65 (49) 14 (47)
Plant intervention
Without plants N(%) 49 (43) 80 (60) 14 (47)
With plants N(%) 65 (57) 54 (40) 16 (53)
Enrollment, informed
consent and allocation to
rooms
n = 472
Blocked
n = 47
Partially blocked
n = 227
Panorama
n = 189
Declined
invitation to
participate (unknown
room number)n = 165
Complete T1-T3 data
n = 114
Complete T1-T3 data
n = 134
Complete T1-T3 data
n = 30
Unknown room
number n = 9
T4 data n = 151 T4 data n = 191 T4 data n = 38
Incomplete
T1-T3 data
but data
available for
T4 n = 62
Incomplete
T1-T3 data
but data
available for
T4 n = 9
Drop
out
n = 5
Drop
out
n = 1
Drop
out
n = 7
Incomplete
T1-T3 data
but data
available for
T4 n = 44
Drop
out
n = 31
Drop
out
n = 31
Drop
out
n = 8
Figure 1. Flow diagram describing the study design and sample available for analysis of change in the health and
well-being outcomes (T1–T3 data) and activity at the center (T4 data).
26 Clinical Rehabilitation 26(1)
Table 2. Means (standard deviations) for the SF-12 physical (P) and mental (M) health scores, emotional state (E) and subjective well-being (W) outcomes for the
three window view conditions from the start of the rehabilitation programme (T1) to the end of the programme (T3)
Panoramic Partially blocked Blocked
Heart Lung Heart Lung Heart Lung
P Men
n¼63
Women
n¼18
Men
n¼20
Women
n¼13
Men
n¼57
Women
n¼12
Men
n¼29
Women
n¼36
Men
n¼12
Women
n¼4
Men
n¼6
Women
n¼8
T1 48.4 (7.83) 42.6 (9.98) 38.6 (9.48) 33.3 (10.0) 45.0 (8.22) 41.6 (11.27) 37.2 (9.28) 38.5 (7.18) 46.5 (9.96) 51.3 (10.99) 33.8 (7.50) 41.1 (9.61)
T2 51.6 (7.82) 46.4 (9.54) 43.3 (9.81) 36.8 (11.71) 50.5 (7.05) 49.8 (8.34) 38.2 (10.12) 40.4 (9.14) 49.7 (8.12) 51.7 (11.12) 34.8 (3.24) 43.6 (7.82)
T3 52.9 (7.79) 46.8 (9.62) 41.1 (10.71) 38.9 (10.71) 52.2 (6.94) 48.1 (7.75) 40.1 (10.43) 40.9 (9.67) 51.2 (8.19) 47.9 (7.78) 42.0 (8.29) 42.1 (8.69)
M
T1 40.2 (9.53) 40.5 (9.65) 34.4 (9.45) 32.4 (14.70) 41.7 (8.44) 45.1 (6.44) 37.0 (8.37) 40.4 (8.29) 43.3 (8.29) 33.5 (4.9) 39.4 (9.28) 35.9 (12.13)
T2 42.2 (8.83) 43.3 (7.87) 42.9 (7.35) 41.1 (8.55) 44.3 (6.90) 40.7 (7.00) 39.0 (9.13) 43.8 (7.46) 41.1 (10.35) 43.9 (6.09) 32.8 (9.19) 41.6 (7.17)
T3 44.1 (7.88) 44.0 (5.75) 40.5 (9.35) 41.4 (8.91) 44.8 (7.88) 49.7 (3.07) 39.4 (9.06) 45.2 (6.28) 41.4 (9.09) 44.6 (0.94) 37.3 (7.49) 41.9 (4.89)
E
T1 4.01 (0.63) 3.89 (0.57) 3.61 (0.63) 3.49 (0.70) 4.09 (0.59) 4.10 (0.37) 3.51 (0.71) 3.78 (0.60) 3.99 (0.61) 4.25 (0.18) 3.53 (0.55) 3.29 (1.03)
T2 4.16 (0.54) 4.05 (0.72) 3.60 (0.71) 3.77 (0.73) 4.22 (0.56) 4.31 (0.33) 3.88 (0.59) 3.98 (0.65) 3.94 (0.65) 4.44 (0.33) 3.48 (0.52) 3.43 (0.91)
T3 4.24 (0.63) 4.19 (0.78) 3.68 (0.76) 3.69 (0.77) 4.28 (0.57) 4.37 (0.23) 3.90 (0.66) 4.16 (0.59) 4.08 (0.67) 4.66 (0.12) 3.72 (0.44) 3.42 (0.51)
W
T1 6.56 (1.58) 6.83 (1.29) 6.28 (1.94) 5.54 (0.88) 6.96 (1.45) 7.08 (1.16) 6.52 (1.62) 6.31 (1.41) 7.00 (2.17) 7.25 (0.50) 6.33 (1.21) 5.38 (0.92)
T2 7.05 (1.31) 7.17 (1.65) 6.55 (1.28) 6.00 (1.22) 7.23 (1.18) 7.41 (1.09) 6.97 (1.82) 6.97 (1.36) 7.67 (1.30) 7.25 (0.96) 6.00 (1.79) 5.50 (1.41)
T3 7.35 (1.39) 7.28 (1.13) 6.90 (1.48) 6.77 (1.54) 7.46 (1.29) 7.75 (1.14) 6.94 (1.36) 7.11 (1.47) 7.67 (1.23) 7.50 (0.58) 6.17 (0.75) 5.38 (1.77)
inclusion of gender and diagnostic group as fac-
tors in the experimental design.
Baseline (T1) data for the measured outcomes
are presented in Table 2. At the start of their
programme, the participants in the different
view groups were similar on all the outcome mea-
sures (all F-values <2.56, all P-values >0.08).
Change in self-reported physical health during
the time at the centre (T1–T3) was modified by
window view and gender together (Tables 2
and 3). Men had similar increases in all view con-
ditions across the three measurement points,
whereas women in the partially and fully blocked
view conditions showed declines during the latter
part of the programme, after initially making
gains (Table 3 and Figure 2 a,b). The change
was also modified by window view and diagnos-
tic group together. Among heart patients, those
with the blocked view showed the least improve-
ment, particularly from T2 to T3. Among lung
patients, those with the panoramic view showed
the greatest improvement from T1 to T2, but
then no further improvement, while the two
other groups continued to improve (Table 3
and Figure 2c,d).
Improvement in mental health during the time
at the centre was also sensitive to the window view,
but the effect was again modified by both gender
and diagnostic group (Tables 2 and 3). A blocked
window view appears to have had a negative effect
on mental health during the period of the rehabil-
itation for men but not for women (Table 3 and
Figure 3). Heart patients with the blocked view
showed the least improvement, while among the
lung patients, those with the panoramic view
showed the greatest improvement from T1 to
T2, but then no further improvement, while the
two other groups continued to improve.
Change in the emotional state and subjective
well-being reports was not modified by window
view. For both variables, all F-values for inter-
actions involving window view were less than
1.07 and all P-values were larger than 0.37.
There were no significant interactions between
window view and the plant intervention condition
for any of the health and well-being measures (all
P-values >0.13). Nor did the window view and
plant intervention together interact with either
gender or diagnostic group (all P-values >0.14).
Patients in the three window view categories
spent about the same number of hours in their
rooms during the daytime (blocked: M¼2.15,
SD ¼1.58; partially blocked: M¼2.12,
SD ¼1.42; panoramic: M¼2.30, SD ¼1.43).
However, more of those who had a panoramic
window view chose to stay in the room when
they wanted to withdraw and be alone in
comparison to those who had a blocked view
(odds ratio (OR) ¼2.32, 95% confidence interval
(CI) ¼1.08–5.01). Those with a partially blocked
view did not have significantly higher odds of
withdrawing to their rooms for solitude than
those in the fully blocked view condition
(OR ¼1.39, 95% CI ¼0.65–2.97). The other
variables in the logistic regression model (age,
gender, diagnostic group, plant intervention
condition) did not contribute to prediction (all
P-values >0.05). Thus, it would appear that,
while spending the same amount of time in the
room during the day, patients with panoramic
views were less likely to spend it in napping.
There were no significant associations between
window view and the measured leisure activities,
including going out for walks and using the cor-
ridor seating areas, seating areas in the reception
and TV rooms (all correlation coefficients <0.09).
In keeping with the health and behavioural
data, patients with a panoramic view to the
surrounding landscape were quite satisfied with
the view, and more so than those who had a
partially blocked view, who were themselves
more satisfied than those with a totally blocked
view (Table 4). Post-hoc tests (Bonferroni)
showed that ratings of each of the view condi-
tions differed significantly from the others
(P<0.03). Satisfaction with the design of the
patient room of itself was high overall and unaf-
fected by the window view (Table 4).
Discussion
The relationship between type of window view
and satisfaction with the view showed a clear
gradient. Patients with a panoramic window
28 Clinical Rehabilitation 26(1)
Table 3. Effects of time at the centre (T1–T3) on self-reported health as moderated by window view, gender and
diagnostic group
FP-value Partial eta
2
SF-12 Physical Time 3view 0.93 0.45 0.01
Time 3view 3gender 2.51 0.04 0.02
Time 3view 3group 2.64 0.03 0.02
SF-12 Mental Time 3view 3.50 0.01 0.03
Time 3view 3gender 5.67 <0.01 0.04
Time 3view 3group 2.76 0.03 0.02
For the physical health outcomes, the Greenhouse–Geisser adjusted degrees of freedom are (3.39, 490.5). For the mental health
outcomes, the Greenhouse–Geisser adjusted degrees of freedom are (3.61, 454.2).
Heart
Time
T1 T2 T3
Mean SF-12 Physical Health
0
34
36
38
40
42
44
46
48
50
52
Panorama
Partially blocked
Blocked
Lung
T1 T2 T3
0
34
36
38
40
42
44
46
48
50
52
Men
(a)
0
34
36
38
40
42
44
46
48
50
52
54
Women
0
34
36
38
40
42
44
46
48
50
52
54
(b)
(c) (d)
Figure 2. Change in mean SF-12 physical health during the rehabilitation programme (T1–T3) as a function of the
window view, for men (a) and women (b) and for heart patients (c) and lung patients (d). Error bars show 61 SE.
Raanaas et al. 29
view to nature were most satisfied, and those
with a blocked view were least satisfied. This
result conforms with a large body of findings
that scenes of nature and natural elements are
preferred over scenes dominated by build-
ings.
23–25
Patients with a panoramic view to nature also
reported using the room as a place to withdraw
to a greater extent. Patients at a rehabilitation
centre are taxed by the demands of a diffi-
cult health situation,
15,16
in addition to the
demands of the rehabilitation programme itself.
Table 4. Satisfaction with window view and room design reported by patients in the three window view conditions
Panoramic Partially blocked Blocked FP-value
Window view 3.34 (0.71) 2.44 (0.99) 1.97 (0.89) 28.04 <0.01
Room design 3.56 (0.58) 3.47 (0.66) 3.50 (0.51) 0.33 0.72
For the window view F-test, df are 2 and 251. For the room design F-test, df are 2 and 252.
Values in the first three columns are means (SDs).
Men
(a)
Mean SF-12 Mental Health
0
34
36
38
40
42
44
46
48
50
52
54
Women
0
34
36
38
40
42
44
46
48
50
52
54
Heart
Time
T1 T2 T3
0
34
36
38
40
42
44
46
48
50
52
Panorama
Partially blocked
Blocked
Lung
T1 T2 T3
0
34
36
38
40
42
44
46
48
50
52
(b)
(c) (d)
Figure 3. Change in mean SF-12 mental health during the rehabilitation programme (T1–T3) as a function of the
window view, for men (a) and women (b) and for heart patients (c) and lung patients (d). Error bars show 61 SE.
30 Clinical Rehabilitation 26(1)
As suggested by Ulrich,
6
by supporting restora-
tion when needed, passive visual contact with the
natural environment may help patients receiving
care to cope better with the demands engendered
by ill-health.
Even though patients at the centre were not in
their bedrooms during the day to the same extent
as in the studies by Ulrich
6
and Park and
Mattson,
26,27
their self-reported health varied
with the type of window view. The effects were
contingent, however, on gender. A blocked
view appears to have negatively affected the
mental health of men but not women during
their stay. On the other hand, a panoramic
view had a more continuously positive effect on
physical health among women. The indoor
plant intervention in the Røros centre also
found gender differences,
17
as did earlier studies
involving indoor nature interventions in other
contexts.
28,29
The effects of the window view over time also
were contingent on diagnostic group. A view
to the surrounding natural landscape appeared
to have a more positive effect on the mental
health of lung patients than of heart patients.
We can only speculate about reasons for the dif-
ferences between the diagnostic groups. The lung
patients generally have more constraints, given
difficulty in breathing and rapid onset of fatigue;
they may therefore need to relax indoors to a
larger extent, and a window view to nature may
therefore have a larger potential for improving
their health. The results of the logistic regres-
sion analysis suggest that the different diagnos-
tic groups did not differ in their use of the
private room, but this does not mean that
they benefitted similarly from the time spent
in the room.
We did not find effects of the window view on
either emotional states or subjective well-being.
This may reflect on the temporal scope of the
measures; the former prompts for emotional
experience at the moment, while the latter asks
the respondent to look back over his or her life.
In contrast, the SF-12 measures place most
emphasis on health as perceived during the
preceding week.
The study has some limitations. Although
we are confident that the allocation procedures
disallowed self-selection, we did not have formal
randomization to view conditions. The outcome
measures were self-reports, and results from
these may not align with objective measures
indicative of the success of the rehabilitation
programme. The analyses here only concern the
experience of the patients while they were at
the centre; long-term health effects cannot be
inferred.
Together with similar studies, the findings
have practical implications. Those designing
health institutions can consider ways to site the
building so that the rooms have pleasing views.
Those responsible for room allocation
can take care to allocate especially vulnerable
persons to rooms with particularly good views.
Clinical messages
.An unobstructed window view to nature
from a private patient room promotes
improvement in self-reported health
during a residential rehabilitation pro-
gramme; however, the strength of the
effect appears to vary as a function of
gender and diagnosis.
.A patient room with an unobstructed
window view to nature may find more use
as a setting to be alone during a challenging
residential rehabilitation programme.
Such withdrawal can help patients cope
emotionally with the demands of their
programme.
Funding
This project was financed with EXTRA funds
from the Norwegian Foundation for Health
and Rehabilitation. It was further supported by
the Norwegian Gardener’s Union, the Bank of
Røros, Tropisk Design, and Primaflor. In addi-
tion to the funders, we thank Tina Bringslimark
for help in planning the study, Tropisk Design
for help in implementing the plant intervention,
Raanaas et al. 31
the staff at the Røros Rehabilitation Centre for their
diligent assistance in conducting the study, and Ellen
Zakariassen for help with data entry.
References
1. Devlin A and Arneill AB. Health care environments and
patient outcomes: a review of the literature. Environ
Behav 2003; 35: 665–694.
2. Rubin HR, Owens AJ and Golden G. An investigation to
determine whether the built environment affects patients’
medical outcomes. Martinez, CA: Center for Health
Design, 1998.
3. Dijkstra K, Pieterse M and Pruyn A. Physical environ-
mental stimuli that turn healthcare facilities into healing
environments through psychologically mediated effects:
systematic review. J Adv Nurs 2006; 56: 166–181.
4. Ulrich RS, Zimring C, Zhu XM, et al. A review of the
research literature on evidence-based healthcare design.
Herd-Health Environ Res Design J 2008; 1: 61–125.
5. Van den Berg AE. Health impact of healing environ-
ments: a review of evidence for benefits of nature, day-
light, fresh air, and quiet in healthcare settings. The
Architecture of Hospitals, Groningen, 2005.
6. Ulrich RS. View through a window may influence recov-
ery from surgery. Science 1984; 224: 420–421.
7. Tang JW and Brown RD. The effect of viewing a land-
scape on physiological health of elderly women.
J Housing Elderly 2005; 19: 187–202.
8. Diette GB, Lechtzin N, Haponik E, Devrotes A and
Rubin HR. Distraction therapy with nature sights and
sounds reduces pain during flexible bronchoscopy: a
complementary approach to routine analgesia. Chest
2003; 123: 941–948.
9. Dijkstra K, Pieterse ME and Pruyn A. Stress-reducing
effects of indoor plants in the built healthcare environ-
ment: the mediating role of perceived attractiveness. Prev
Med 2008; 47: 279–283.
10. Ulrich RS, Simons RF and Miles MA. Effects of envi-
ronmental simulations and television on blood donor
stress. J Archit Plan Res 2003; 20: 38–47.
11. Hartig T, Evans GW, Jamner LD, Davis DS and Ga
¨rling
T. Tracking restoration in natural and urban field set-
tings. J Environ Psychol 2003; 23: 109–123.
12. Ulrich RS, Simons RF, Losito BD, Fiorito E, Miles MA
and Zelson M. Stress recovery during exposure to natu-
ral and urban environments. J Environ Psychol 1991; 11:
201–230.
13. Gerlach-Spriggs N, Kaufman RE and Warner Jr SB.
Restorative gardens: the healing landscape. New Haven,
CT: Yale University Press, 1998.
14. Cooper Marcus C. Acute care general hospitals: typol-
ogy of outdoor spaces. In: Marcus CC, Barnes M (eds)
Healing gardens: therapeutic benefits and design recom-
mendations. New York: John Wiley & Sons, 1999,
pp.115–156.
15. Normann T, Sandvin JT and Thommesen H. A holistic
approach to rehabilitation. Oslo: Kommuneforlaget, 2004.
16. Becker G. Disrupted lives: how people create meaning in a
chaotic world. Berkeley, CA: University of California
Press, 1997.
17. Raanaas RK, Patil GG and Hartig T. Effects of an indoor
foliage plant intervention on patient well-being during
a residential rehabilitation program. HortScience 2010;
45: 1–6.
18. Ware JE, Kosinski M and Keller SD. A 12-item short-
form health survey: construction of scales and prelimi-
nary tests of reliability and validity. Med Care 1996; 34:
220–233.
19. Ware JE, Kosinski M, Turner-Bowker DM and Gandek
B. How to score version 2 of the SF-12 Health Survey.
(With a supplement documenting version 1). Boston,
MA: Health Assessment Lab, 2005.
20. Abdel-Khalek AM. Measuring happiness with a single-
item scale. Soc Behav Personal 2006; 34: 139–149.
21. Knez I and Hygge S. The circumplex structure of affect: a
Swedish version. Scand J Psychol 2001; 42: 389–398.
22. Tabachnick BG and Fidell LS. Using multivariate statis-
tics, fifth edition. Boston, MA: Pearson/Allyn and
Bacon, 2007.
23. Hartig T. Nature experience in transactional perspective.
Landscape Urban Plan 1993; 25: 17–36.
24. Kaplan R and Kaplan S. The experience of nature: a
psychological perspective. Cambridge: Cambridge
University Press, 1989.
25. Ulrich RS. Aesthetic and affective response to natural
environment. In: Altman I, Wohlwill JF (eds) Human
behavior and environment. New York: Plenum Press,
1983, pp. 85–125.
26. Park SH and Mattson RH. Effects of flowering and
foliage plants in hospital rooms on patients recovering
from abdominal surgery. HortTechnology 2008; 18:
563–568.
27. Park SH and Mattson RH. Therapeutic influences of
plants in hospital rooms on surgical recovery.
HortScience 2009; 44: 102–105.
28. Campbell DE. Interior office design and visitor response.
J Appl Psychol 1979; 64: 648–653.
29. Dravigne A, Waliczek TM, Lineberger RD and Zajicek
JM. The effect of live plants and window views of green
spaces on employee perceptions of job satisfaction.
HortScience 2008; 43: 183–187.
32 Clinical Rehabilitation 26(1)
