Content uploaded by Judith Redi
Author content
All content in this area was uploaded by Judith Redi on Jul 30, 2015
Content may be subject to copyright.
RESEARCH ARTICLE
Lighting to Make You Feel Better: Improving
the Mood of Elderly People with Affective
Ambiences
Andre Kuijsters
1
*, Judith Redi
2
, Boris de Ruyter
3,4
, Ingrid Heynderickx
1
1 Human Technology Interaction Group, Industrial Engineering and Innovation Science, Eindhoven
University of Technology, Eindhoven, the Netherlands, 2 Multimedia Computing Group, Intelligent Systems,
Delft University of Technology, Delft, The Netherlands, 3 Philips Research, Eindhoven, The Netherlands,
4 Human Interaction with Intelligent Systems, Social sciences, Radbout University, Nijmegen, The
Netherlands
* andrekuijsters@gmail.com
Abstract
Current lighting technologies extend the options for changing the appearance of rooms and
closed spaces, as such creating amb iences with an affective meaning. Using intelligence,
these ambience s may ins tantly be adapted to the needs of the room’s occupant(s), possibly
improving their well-being. We hypothesized that ambiences with a clearly recognizable,
positive affective meaning could be used to effectively mitigate negative mood in elderly.
After inducing a sad mood with a short movie one group of elderly was immersed in a posi-
tive high arousing (i.e., activating) ambience, and another group in a neutral ambience. Sim-
ilarly, after inducing anxiety with a short movie one group of elderly was immersed in a
pleasant low arousing (i.e., cozy) ambience, and another group in a neutral ambience. We
monitored the evolution of the mood of the four groups of elderly over a period of ten min-
utes after the mood induction, with both self-reported mood measurements (every 2 min-
utes) and constant measurements of the skin conductance response (SCR) and
electrocardiography (ECG). In line with our hypothesis we found that th e activating ambi-
ence was physiologically more arousing than the neutral ambience. The cozy ambience
was more effective in calming anxious elderly than the neutral ambience, as reflected by
both the self-reported and physiological measurements.
Introduction
Ambient intelligence research is devoted to designing systems that, embedded in our everyday
environments, can help in improving the quality of our lives [1]. Among the various facets of
ambient intelligence, a particularly interesting one is the so-called “empathic technology”: sys-
tems able to detect the affective state of a person, and react on it by providing the necessary
care [2]. This type of systems is especially interesting for improving the life of vulnerable peo-
ple, e.g., children, people with disabilities, or elderly. In our research, we focus on creating
PLOS ONE | DOI:10.1371/journal.pone.0132732 July 20, 2015 1/22
a11111
OPEN ACCESS
Citation: Kuijsters A, Redi J, de Ruyter B,
Heynderickx I (2015) Lighting to Make You Feel
Better: Improving the Mood of Elderly People with
Affective Ambiences. PLoS ONE 10(7): e0132732.
doi:10.1371/journal.pone.0132732
Editor: Marina A. Pavlova, University of Tuebingen
Medical School, GERMANY
Received: February 10, 2015
Accepted: June 17, 2015
Published: July 20, 2015
Copyright: © 2015 Kuijsters et al. This is an open
access article distributed under the terms of the
Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any
medium, provided the original author and source are
credited.
Data Availability Statement: Data are available from
the Open Science Framework with the following DOI:
10.17605/OSF.IO/ED97R.
Funding: This research is part of the ACE project,
which is a collaboration between Philips Research,
Eindhoven University of Technology, and Delft
University of Technology, funded by PointOne of
Agentschap NL [PNU 10C11]. Website: http://www.
rvo.nl/. Co-author BdR is employed by Philips
Research. Philips Research provided support in the
form of salary for author BdR, but did not have any
additional role in the study design, data collection and
analysis, decision to publish, or preparation of the
empathic technology for the elderly, going through a particularly difficult time of their life: the
re-location to care centers.
Elderly in care centers often experience negative mood states [3,4]. They can be anxious
because they realize they are in the last stage of their life, or sad because they miss their family,
friends or pets [4]. Countering such negative moods is important, given that positive moods
have been linked to increased health and well-being [5], and more prosocial behavior [6].
Human intervention (e.g., from care takers) is possibly the most effective way to improve
elderly’s mood; however, often due to limited resources, human intervention is not always pos-
sible. As a consequence, equipping care centers with intelligent technology, able to detect nega-
tive moods and drive them towards more pleasant ones, may bring great added value to the
well-being of care center inhabitants.
One way to achieve this is to use intelligent lighting solutions to create pleasant ambiences
in care center rooms. The vast growth in the amount of LED sources, their miniaturization and
the possibility to integrate sensors and electronics has created a wealth of options to automati-
cally generate light of any color and intensity at any place or time with very thin, unobtrusive
luminaries, making it possible to re-configure light according to a promptly detected negative
mood of a room’s occupant. Light and color in general have been repeatedly shown to be able
to influence people’s mood, both at a biological and psychological level. From a biological
point of view, (short wavelength) light can influence the circadian rhythm, and as such the
mood [7]. At a psychological level, (colored) lighting has been shown to carry meaning through
associations, thereby influencing affective states. For example, employees working in colorful
offices tend to experience more positive moods than employees sitting in neutral or colorless
offices [8]. Different hues are also commonly associated with arousal levels: blue, for example,
is deemed to be calming, whereas red is considered to be more energizing [9,10].
In addition, previous literature has shown that it is possible to use lighting to create ambi-
ences whose affective connotation (also referred to as atmosphere [11], e.g. cozy or activating)
can be clearly recognized by people [12,13]. This perceived affect may in turn impact the mood
of the room's occupant. It is unknown, however, whether the exposure of a person to an affec-
tive ambience has the power to steer the person’s mood towards an affective state that matches
the ambience connotation.
In this study, we hypothesize that affective ambiences, created with lighting, contribute to
improve people's mood. We envision an Adaptive Ambience Creation platform that can auto-
matically adapt the ambience in a care centre room to the needs of the elderly occupying it, and
in this paper, we investigate its effectiveness. We regard mood as two-dimensional construct
with valence and arousal as bipolar dimensions [14]. Valence (pleasure) summarizes how well
one is doing, whereas arousal (activation) refers to a sense of mobilization or energy. We focus
on countering, at a psychological rather than biological level, two negative moods commonly
encountered in elderly just relocated in care centres, i.e. anxiety (negative valence and high
arousal) and sadness (negative valenc e and low arousal) [3,4]. To do so, we resort to two affec-
tive ambiences, with an associated (and found to be recognized by elderly [12]) relaxing and
activating atmosphere, respectively. These ambiences are created with the combination of gen-
eral white lighting, varying in illuminance and color temperature, and colored accent lighting.
We expect an ambience recognized as relaxing (more specifically, cozy) to lower arousal and
improve the valence on an anxious elderly. Similarly, we expect an ambience recognized as acti-
vating to increase both arousal and valence of an elderly in a sad or gloomy mood.
This paper presents an experiment through which we provide evidence that carefully
designed lighting ambiences with a clearly recognized positive affective connotation can help
in improving the mood of elderly in the direction of the recognized affective connotation. To
show this, we first induce a negative mood (i.e., either anxiety or sadness) in elderly
Lighting to Make You Feel Better
PLOS ONE | DOI:10.1371/journal.pone.0132732 July 20, 2015 2/22
manuscript. The specific role of this author is
articulated in the ‘author contributions’ section.
Competing Interests: BdR is an employee of Philips
Research, whose company partly funded this study
through the ACE project. There are no patents,
products in development or marketed products to
declare. This does not alter the authors' adherence to
all the PLOS ONE policies on sharing data and
materials.
participants. We then investigate whether a high arousing pleasant ambience (referred to as
activating) can counter sadness more than a neutral ambience. Equivalently, we measure
whether a low arousing pleasant ambience (referred to as cozy) can mitigate anxiety more than
a neutral ambience. Specifically, we formulate these two hypotheses:
H-
1.An
activating ambience is more effective in increasing both pleasure and arousal in elderly people
that are in a sad mood than a neutral ambience.
H-
2.A
cozy ambience is more effective in both increasing pleasure and reducing arousal in elderly
people that are in an anxious mood than a neutral ambience.
Both hypotheses are visu alized in Fig 1, and were tested by using self-reported and psycho-
physiological measures.
Related Work
The relationship between lighting and affect has been studied from different perspectives and
is largely documented; results are, however, not always coherent. We review results on the bio-
logical and psychological influence of light ing and color on mood. The biological (or non-
visual) effects of light can be divided in indirect (via circadian photo entrainment) and direct
effects on alertness and sleepiness [7,15]. The effects of light can be measured directly on mela-
tonin or cortisol levels, but also via self-reported mood and physiological mood measures [15].
The psychological effects of lighting on mood are mediated by, amongst others, associations
and preferences [10]. These effects are often measured with self-reported mood, but also with
(psycho)physiological mood measures and via mood related behavior. In the last section we
briefly discuss the existing knowledge on the affective connotation (i.e. atmosphere perception)
of lighting ambiences.
Influence of lighting on mood at a biological level
The indirect effects of lighting on mood are well documented [7,16,17]. Night time release of
melatonin—a pineal hormone that promotes sleepiness and lowers the body temperature—can
be phase advanced or delayed with the right exposure to bright light [18], thereby influencing
the circadian rhythm of humans. Disruptions to circadian rhythms (e.g. due to jetlag, shift
work, night time light exposure) can cause sleep and mood disorders [19–21]. Seasonal affec-
tive disorders (characterized by depression symptoms in the winter) are also conside red to be
the result of a disturbance of the circadian rhythm, caused by the failure to adapt to the shift in
day length as a result of seasonal changes [19]. Bright light therapy of 1500 lx or more at eye
level, has been shown to significantly reduce depression symptoms for people with seasonal
mood disorders [21,22] or sleep disorders [21].
The wavelength of the light is also important; melatonin is especially suppressed by short
wavelength light [23 ,24]. Thapan and colleques [24] revealed that melatonin was largely sup-
pressed by blue light, slightly by green light and hardly or not at all by red light. This finding
suggests that, at a biological level, blue light is more arousing than green light, and both are
more arousing than red light.
Recent research also suggests direct effects of lighting on alertness and sleepiness [7,15,25].
Using different light spectra, studies reported the superiority of short wavelength light over
long wavelength light in increasing alertness [7,15].
Lighting to Make You Feel Better
PLOS ONE | DOI:10.1371/journal.pone.0132732 July 20, 2015 3/22
Influence of lighting on mood at a psychological level
Along with biological effects, (colored) lighting has also strong psychological effects on the mood
of people. Yoto and colleagues [26] found participants rating red colored cards as more arousing
than blue colored cards. Kaiser [10] postulated that the arousing property of red may be medi-
ated by associations with past experiences; red is often associated with danger and is often used
as warning sign, while blue is associated with the sea or clear sky. In line with this argumentation,
Küller and colleagues [9] revealed that a room painted red decreased EEG alpha activity of the
participants more than a blue room, indicating an increased arousal in the red room. Earlier
studies on the effect of colored lighting on the mood of people revealed similar results. Ali [27]
revealed that red light suppressed EEG alpha activity more than blue light, indicating that red
light was more arousing than blue light. Wilson [28] found that red light increased skin
Fig 1. Schematic representation of the hypotheses formulated in this study. (H1) an activating ambience is more effective in increasing both pleasure
and arousal in elderly that are in a sad mood than a neutral ambience, and (H2) a cozy ambience is more effective in both increasing pleasure and reducing
arousal in elderly that are in an anxious mood than a neutral ambience.
doi:10.1371/journal.pone.0132732.g001
Lighting to Make You Feel Better
PLOS ONE | DOI:10.1371/journal.pone.0132732 July 20, 2015 4/22
conductance compared to green light. Jacobs and Hustmeyer [29] found that red light projected
on a screen was more arousing than green light, and green light more arousing than yellow and
blue light, as measured by an increased skin conductance. Gerard [30] found that red light was
more arousing than blue light, a conclusion that was deduced from an increased augmented sys-
tolic blood pressure, skin conductance, respiration rate, and cortical activation. He reported that
red light yielded a variety of unpleasant associations involving blood, fire, danger as well as
aggression, while blue light yielded associations with friendliness, blue skies, blue water of a lake
or an ocean. These associations with colors might explain Gerard’s findings as he gave his partici-
pants the instruction: “If the colors bring certain thoughts and associations to your mind, allow
yourself to have these thoughts, whatever they are” (p. 51) [30]. In contrast to the above findings,
blue colored cards were found to be more arousing than red colored cards, as evidenced from
subjective evaluations [31] and a lower electroencephalography (EEG) alpha and theta activity
[26]. Nourse and Welch [32] found that test participants’ skin conductance response indicated
more arousal when exposed to purple light compared to green light.
With respect to valence, it is often reported that short wavelength colors (such as blue,
green) evoke more pleasant feelings than long wavelength colors (such as orange, red)
[31,33,34]. This effect was found for office workers [34] and in shopping environments [33]. In
contrast to these findings, Küller and colleagues [9] found that participants felt significantly
more pleasant in a red room than in a blue room; this, however, was observed only in one of
two reported experiments. Stone [35] found that the context was very important; positive
mood increased in a blue open-plan study area compared to a red open-plan study area; how-
ever, opposite results were found for a private setting. In another study, Küller and colleagues
[8] found that the presence of any color could positively influence people’s mood in work envi-
ronments, although the affective response to colored walls was also depending on the context.
Not only the context but also the lightness and saturation of the color(ed lighting) was
found to be very important for its influ ence on mood. Kaiser [10] argued that the green and
red stimuli in the experiment of Wilson [28] were not equal in lightness and saturation level,
and that the saturation differences might account for the findings. Also in the experiment of
Nourse and Welch [32] the purple and green stimuli were not equal in saturation; the purple
stimulus was more saturated. The higher saturation might be a better explanation for the
higher skin conductance found in their experiment than the difference in hue. In line with this
argumentation, Valdez and Mehrabian [31] revealed that the affective evaluation of colors was
mainly based on saturation and lightness and less on hue. Similar results were found by Mikel-
lides [36]; the saturation rather than the hue determined the arousing effect of a color. The
lightness of a color was shown to increase arousal [15] and valence ratings [31]. Also saturation
increased the valence ratings [31]; however, when the colors were too saturated, the effect
became undesired. Kwallek and colleagues [34] found that highly saturated colors in an office
induced more depression, confusion and anger with males.
The psychological effects of white light on mood have been studied extensively as well,
revealing though mixed results. Some studies found no significant effect of white light on
mood [37,38]. Although Baron and colleagues [37] did not find a direct effect of lighting set-
tings on mood, they found several consistent effects of lighting on indirect measurements of
mood. Warmer light (3000K) was found to enhance the willingness to help in a next experi-
ment and to resolve conflicts as compared to cooler white light (4200K). Lower illuminance
(150lx compared to 1500lx) was found to enhance ratings for an imaginary employee and to
increase the amount of time subjects were willing to spend voluntarily on a next experiment.
Knez and colleagues [39] argued that the exposure time in these studies was not long enough
to induce mood changes. They performed several experiments [39–41] with a longer exposure
time (i.e. 90 till 120 minutes). In their studies, females’ negative mood decreased while working
Lighting to Make You Feel Better
PLOS ONE | DOI:10.1371/journal.pone.0132732 July 20, 2015 5/22
under warmer light conditions (3000K) and increased under cooler light conditions (4000K),
while an opposite effect was found for males; no significant effect was found for positive mood
[39]. In a later study Knez and Enmarker [40] found an opposite result; females’ negative
mood increased more under warmer white light (3000K), while males’ negative mood
increased more in cooler white light (4000K). Females preserved their positive mood best
under cooler white light (4000K), while males under warmer white light (3000K).
Finally, Küller and colleagues [8] investigated the effect of (indoor and outdoor) lighting on
the mood of people working indoors in real work environments. Effects were measured in dif-
ferent seasons and different countries with different latitudes. Although they found no effect of
illumination on the mood measurements, they found a strong correlation between mood and
the appraisal of the lighting. If the lighting was judged right, self-reported mood was high,
while mood decreased when the lighting was judged to be too dim or too bright. Similar results
are reported by Veitch and colleagues [42]; people that judged the lighting in their room of
higher quality reported a better mood.
Atmosphere perception
The results described above reveal the high potency of lighting to reduce negative mood. How-
ever, drawing clear conclusions in terms of which light characteristic to use for a more positive
mood is difficult as mixed outcomes are found, due to, amongst others., the discrepancy
between psychological and biological influen ces, the context, and the interactions between the
different lighting characteristics. In addition, it should be considered that people’s mood is not
only affected by lighting, but also by other environmenta l factors (e.g., temperature) and non-
environmental factors (e.g., internal state and personal preferences). In designing ambiences
that can effectively improve the mood of elderly, it is therefore useful to first evaluate the per-
ception of the ambience, i.e., the atmosphere. We argue that atmosphere perception mediates
the psychological influence of lighting on mood.
Atmosphere is not an affective state in itself, but an affective evaluation of the environment
[11]. Past studies have shown that atmosphere can be characterized along four dimensions,
namely coziness, liveliness, detachment and tenseness [11], of which liveliness and coziness are
the most important ones [11,43]. Alternatively, atmosphere may be evaluated in terms of
valence and arousal [44,45]. The relation between both characterizations of atmosphere is that
coziness was found to be positively correlated with positive valence and negatively correlated
with arousal, while liveliness was found to be positively correlated with both valence and
arousal [12]. Extensive kn owledge has been built up about the influence of lighting characteris-
tics on atmosphere percept ion [12,13,43,46 ]. The combination of general white lighting with
colored accent lighting is known to be able to generate cozy, activating and exciting atmo-
spheres in a room for younger people [12,47]. Despite the fact that atmosphere perception is
different between younger and elderly people, it was possible to retrieve ambiences with a clear
affective connotation for elderly too [12]. Nevertheless, it is so far unclear whether clearly rec-
ognized atmospheres have the power to influence the mood of people immersed in them, con-
ditioning their mood towards matching the affective value of the atmosphere. This study is set
out to verify whether that is the case.
Method
To verify whether positive atmospheres could counter negative moods, we identified two nega-
tive moods commonly experienced by elderly in care centers [48], namely anxiety (i.e., negative
valence and high arousal) and sadness (i.e., negative valence and low arousal), and two atmo-
spheres with an opposite affective connotation, namely cozy (i.e., positive valence and low
Lighting to Make You Feel Better
PLOS ONE | DOI:10.1371/journal.pone.0132732 July 20, 2015 6/22
arousal) and activating (i.e., positive valence and high arousal) (see also Fig 1). To test the effec-
tiveness of the cozy atmosphere we first induced an anxious mood. Then we exposed one
group of participants to the cozy atmosphere and the other to a neutral one. We then moni-
tored the mood of the participants in both groups for a period of 10 minutes. Our hypothesis
was that we would find more positive valence and lower arousal (measured both self-reported
and physiologically) for participants exposed to the cozy atmosphere than to the neutral one.
The effectiveness of the activating atmosphere was tested in a similar way; inducing first sad-
ness and then exposing one group to the neutral atmosphere and the other group to the activat-
ing one. Next we describe the details of the experimental setup.
Participants
Thirty-eight older people, with an equal number of males and females, participated in the
experiment. Their age varied between 66 and 94 years (M = 78.8 and SD = 6.8). They were
recruited from the independent living unit of the Vitalis WoonZorg Groe p in Eindhoven, The
Netherlands. At the end of the experiment they were given a token of appreciation for their
participation. This study was approved by the institutional review board of the Eindhoven Uni-
versity of Technology, following the Code of Ethics of the Dutch Institute for Psychologists,
and by the ethical committee (ICBE) of Philips Research, following th e Declaration of Helsinki,
in 2013. All participants gave written informed consent.
Experimental room
The experimental room was located at the Philips ExperienceLab and had dimensions of 7.4 x
4.2 x 2.8 meters. The walls were white, the ceiling had an off-white (i.e., ivory white) color, and
the floor consisted of dark grey carpet tiles. The windows were blinded, in order to prevent influ-
ences of natural light during the experiments, and were covered with low chromatic curtains.
The room was furnished as a living room with a white sofa and a white chair around a black cof-
fee table. An off-white carpet was placed underneath the coffee table. Against the wall facing the
sofa stood a black television cabinet with a black 42” television on it. Finally, a black dinner table
with four off-white chairs were placed against the wall on the left-hand side of the sofa.
An overview of the installed luminaires is given in Fig 2. Functional white lighting was pro-
vided by two cylindrical floor lights consisting of four fluorescent lamps each; two lamps with a
warm white color temperature (CT) of 2700K (Philips Master TL5 HE 28W/827) and two
lamps with a cold white of 6500K (Philips Master TL5 HE 28W/865 lamps). Additional white
light was provided by six pairs of halogen spot lights; each pair consisted of one spot with a
warm white CT of 3000K (Philips HR Dich roic 50W GU5.3 12V 36D) and one spot with a cool
white CT of 4700K (Philips Diamondline 50W GU5.3 12V 36D 1CT). Decorative lighting was
generated by three Philips Living Color lamps. Two were placed on each side of the television
cabinet and one in the upper left corner. A table light consisting of red, green and blue LED
strips was mounted underneath the coffee table and illuminated the floor locally. Finally, a
Gemini lamp, consisting, on the one hand, of red, green and blue LEDs for illuminating the
ceiling, and, on the other hand, of white LEDs for illumination downwards, was mounted
above the dinn er table.
Ambiences
We used ambiences (pictured in Fig 3) for which we had confirmation that elderly could clearly
recognize their atmosphere. These ambiences were designed and validated in a previous study
[12]. The cozy ambience consisted of functional lighting at a lower illuminance (i.e., 120 lx)
and lower CT (i.e., 2700 K), to which we added orange colored accent lighting. The Gemini
Lighting to Make You Feel Better
PLOS ONE | DOI:10.1371/journal.pone.0132732 July 20, 2015 7/22
lamp did not illuminate the dinner table in the cozy ambience. The cozy ambience was per-
ceived as a pleasant low arousing ambience, as reflected by high coziness scores in the atmo-
sphere questionnaire [12]. Coziness was found to be positively correlated with positive valence
and negatively correlated with arousal [12]. The activating ambience consisted of functional
lighting at a higher illuminance (i.e., 325 lx) and a higher CT (i.e., 4000 K), to which we added
cyan colored accent lighting [12]. The activating ambience was perceived as a pleasant high
arousing ambience, as reflected by high liveliness scores in the atmosphere questionnaire. Live-
liness was found to be positively correlated with both valence and arousal [12]. The neutral
ambience consisted of functional white lighting with an intermediate illuminance (i.e., 150 lx)
and intermediate CT (i.e., 3400 K), and here we added white accent lighting. Pilot tests revealed
that this ambience was scored neutral on both coziness and liveliness in the atmosphere ques-
tionnaire [11]. The reported illuminance and CT values were all measured vertically at the
position of the participant’s eyes. More details about the design and validation of the cozy and
activating ambiences can be found in [12].
Fig 2. An overview of the installed luminaires in the experimental room.
doi:10.1371/journal.pone.0132732.g002
Fig 3. An impression of the activating ambience (left) and cozy ambience (right).
doi:10.1371/journal.pone.0132732.g003
Lighting to Make You Feel Better
PLOS ONE | DOI:10.1371/journal.pone.0132732 July 20, 2015 8/22
Mood induction
Different procedures for inducing mood in a controlled way exist (for reviews see [49–51]),
and include listening to music, watching movies, watching affective slides, imagining scenes,
and recalling past life events. For our study we chose a mood induction procedure based on
viewing movie excerpts. This choic e was based on the fact that induction with visual stimuli is
less sensitive to demand characteristics [49], most effective in inducing the targeting negative
mood states [49,51], and standardized material for the procedure is publicly available [52].
Two short film segments were selected from the database of Gross and Levenson [52]. For
inducing sadness, the scene at which a boy cries at his father’s death (2’51”) from the movie
‘The Champ’ was selected [53]. For inducing anxiety, the basement chase scene (3’29”) from
the movie ‘The silence of the lambs’ was selected [54].
Mood measurements
The participants assessed their actual mood state with the pleasure and arousal scale of the
Self-Assessment Manikin (SAM) [55]. Nine-point visual scales were used to assess the subjec-
tive level of pleasure and arousal of the participants.
During the trials we also monitored the electrodermal activity (via the skin conductance
response, SCR) and the cardiac response (via the ECG signal) of the participants. Heart Rate
(HR) and Heart Rate Variability (HRV) were computed from the ECG signal. Skin conduc-
tance response and HR increase are mainly linked to increased arousal [56–59]. Increased
HRV is an indicator of an increased parasympathetic activity (i.e., resting state) and/or
decreased sympathetic activity (i.e., active state). An increase in the high frequency component
of HRV has been linked to an increase in parasympathetic activity.
The physiological responses were measured with the Nexus-10 system (Mind Media BV,
Roermond, The Netherlands). The device was controlled with the Biotrace software suite (Mind
Media BV, Roermond, The Netherlands) via a computer. The skin conductance response was
recorded with the Nexus-10 SC/GSR sensors. The sensors were positioned to the upper phalan-
ges of the index and middle finger of the non-dominant hand. For the ECG recordings, pre-gelled
silver chloride electrodes were placed at the inside of the wrists of both arms, and a third elec-
trode, functioning as ground sensor, was placed at the upper arm. The saving rate and resolution
of the physiological measurements were: 32 Hz and 0.001 μS, respectively for SCR, and 2048 Hz
and 1 μV, respectively, for ECG. Heart Rate (HR) and Heart Rate Variability (HRV) were com-
puted offline from the ECG recordings by analyzing the variability in the intervals between sinus
rhythm heart beats (i.e., R–R intervals), following the guidelines of the task force of the European
Society of Cardiology and the North American Society of Pacing and Electrophysiology.[60]Cus-
tom software was written in MATLAB for the R-wave detection and the complete signal was
carefully visually inspected for false or undetected R-waves, ectopic beat errors, and movement
artefacts, which were then manually corrected. HR (beats per minute) was calculated from the
inter-beat-intervals. Time and frequency domain HRV measures were analyzed using KARDIA
[61] in MATLAB. The RMSSD (i.e., the square root of the mean squared differences of successive
R-R intervals) was calculated as indicator of short term HRV. The High Frequency component
(HF-HRV; 0.15–0.40 Hz) of HRV was calculated in the frequency domain with Fast Fourier
Transform analysis (resample rate = 2 Hz, FFT window length = 512). The HF-HRV is a specific
indicator of parasympathetic activity and is expressed in absolute values of power (in ms
2
).
Procedure
The participants were welcomed in the experimental room and asked to sit on the sofa three
meters in front of the 42 inch television screen. Each participant conducted two sessions; one
Lighting to Make You Feel Better
PLOS ONE | DOI:10.1371/journal.pone.0132732 July 20, 2015 9/22
session with a sad mood induction and one session with an anxious mood induction. The
order of the mood induction was randomized. After the mood induction the influence of the
ambience was investigated between subjects (Activating versus Neutral ambience after sad
mood induction; Cozy versus Neutral ambience after the anxious mood induction). The order
of the ambiences was also randomized resulting in four groups of experimental conditions (as
also explained in Table 1). We counterbalanced male and female participants in each group.
The participants entered the experimental room with a neutral lighting. After signing the
informed consent the sensors for the physiological measurements were applied. Hereafter a fif-
teen minutes waiting period started to ensure that the preparations of the physiological mea-
surements did not influence the baseline mood measurements. During the first five minutes of
this waiting period, the experimenter explained the setup and the questionnaire scales to the
participants. After the fifteen minutes waiting period, the participants verbally indicated their
subjective mood on that moment based on the SAM scales. The lighting was then dimmed to a
low level (i.e., 40 lx) and the mood inducing short movie was displayed on the television screen.
After the mood induction, the second SAM measure was collected. Hereafter the lighting was
changed to a neutral, cozy or activating setting, depending in which group and session the par-
ticipant was in. From that moment, the participants indicated their actual mood every two
minutes with the SAM scale. After the seventh mood measurement (i.e., after 10 minutes), a
short break was incorporated. The second session followed a procedure similar to the first ses-
sion with only a shorter relaxing period for the baseline measurement (i.e., 5 minutes). After
the experiment, participants were thanked for their participation and debriefed. They received
a coupon of 20 euros for their participation. The whole experimental procedure is schemati-
cally shown in Table 2.
Results
Equipment and recording errors led to missing data for the physiological measurements. The
sample sizes for each of the dependent measures are depicted in Table 3. We found no carry-
ing-over effects between the first and second session; the baseline scores from the first session
were not significantly different from the baseline scores of the second session for all the mea-
sures, as tested with paired-samples t-tests.
Effectiveness of mood induction
To investigate the influence of the sad and anxious mood induction on the affective ratings and
physiological measures, we tested a series of general linear models with gender and ambience
group as between-subjects factors and mood induction (pre/post) as the within-subjects factor.
For the physiological measures we compared the median values of the last 120 seconds of the
baseline with the median values of last 120 seconds of the mood induction.
Table 1. Distribution of participants over the experimental conditions; in the ‘Session1’ and ‘Session
2’ columns the first adjective refers to the induced mood, and the second adjective to the atmosphere
used; for example, "Sad—Neutral" refers to inducing a sad mood, and evaluating its reduction in a
neutral ambience.
Group Age Session 1 Session 2
1 (n = 9; 5 female) M = 76.8, SD = 8.8 Sad—Neutral Anxious—Cozy
2 (n = 9; 4 female) M = 78.5, SD = 6.9 Sad—Activating Anxious—Neutral
3 (n = 10; 5 female) M = 80.5, SD = 8.2 Anxious—Cozy Sad—Neutral
4 (n = 10; 5 female) M = 79.2, SD = 7.7 Anxious—Neutral Sad—Activating
doi:10.1371/journal.pone.0132732.t001
Lighting to Make You Feel Better
PLOS ONE | DOI:10.1371/journal.pone.0132732 July 20, 2015 10 / 22
Fig 4 presents the subjective pleasure (left) and arousal (right) scores before and after the
mood induction for both the sad and anxious movie. For the sad movie we found that the plea-
sure ratings were significantly reduced after the mood induction (F(1, 34) = 120.92, p < .001,
η
2
= .78) as intended. Also the effect of gender reached significance for pleasure (F(1, 34) =
6.33, p = .017, η
2
= .16); further analyses revealed that males scored higher on pleasure at base-
line (M = 7.11) and after the sad movie (M = 4.21) than females (at baseline M = 6.26, and after
the sad movie M = 2.95). In contrast with our expectations the SAM arousal ratings were sig-
nificantly increased by the mood induction (F(1, 34) = 37.27, p < .001, η
2
= .52). We found
also here a gender effect (F(1, 34) = 8.27, p = .007, η
2
= .20). Further analysis revealed that
females scored higher on arousal at baseline (M = 3.21) and after the sad movie (M = 5.53)
than males (at baseline M = 2.11, and after the sad movie M = 4.32). For the physiological mea-
sures a significant effect of the sad mood induction on SCR and HR was found. In line with the
SAM arousal ratings, also the SCR significantly increased during the mood induction (F(1, 32)
= 4.44, p = .043, η
2
= .12) from M = 4.09 at baseline to M = 4.34 during mood induction. Also
the HR increased significantly (F(1, 31) = 4.89, p = .034, η
2
= .14) from M = 66.58 at baseline to
M = 67.40 during mood induction. Furthermore a gender x ambience group interaction effect
was found on HR (F(1, 31) = 9.44, p = .004, η
2
= .23); further analysis revealed that the females
in the neutral ambience group had a higher HR (M = 74.1) than the male participants
(M = 60.6), whereas a reversed effect was found for the activating ambience group (females
M = 63.2, and males M = 70.9).
Table 2. Schematic overview of the experimental procedure.
Session 1 Session 2
Neutral lighting Preparations physiological
measurements
Break
waiting period (15 min) waiting period (5 min)
Baseline mood (SAM) Baseline mood (SAM)
Film lighting Mood induction (Sad film) Mood induction (Anxious
film)
Mood after MI (SAM) Mood after MI (SAM)
Cosy/activating or neutral
lighting
waiting period (2 min) waiting period (2 min)
2 min (SAM) 2 min (SAM)
waiting period (2 min) waiting period (2 min)
4 min (SAM) 4 min (SAM)
waiting period (2 min) waiting period (2 min)
6 min (SAM) 6 min (SAM)
waiting period (2 min) waiting period (2 min)
8 min (SAM) 8 min (SAM)
waiting period (2 min) waiting period (2 min)
10 min (SAM) 10 min (SAM)
doi:10.1371/journal.pone.0132732.t002
Table 3. Sample sizes for the different dependent measures for each ambience group.
Ambience group Pleasure and arousal ratings Skin conductance response Heart rate and Heart rate variability
Sad movie-> neutral ambience n = 19 n = 18 n = 17
Sad movie-> activating ambience n = 19 n = 18 n = 18
Anxious movie-> neutral ambience n = 19 n = 18 n = 17
Anxious movie-> cozy ambience n = 19 n = 19 n = 16
doi:10.1371/journal.pone.0132732.t003
Lighting to Make You Feel Better
PLOS ONE | DOI:10.1371/journal.pone.0132732 July 20, 2015 11 / 22
After the anxious mood induction the (self-reported) pleasure ratings were significantly
reduced as expected (F(1, 34) = 76.38, p < .001, η
2
= .69). A significant effect of ambience
group on pleasure was found (F(1, 34) = 6.17, p = .018, η
2
= .15). As can be seen in Fig 4, the
elderly in the cozy ambience group scored lower on pleasure at baseline and after the mood
induction. Also the effect of gender reached significance (F(1, 34) = 9.53, p = .004, η
2
= .22);
further analysis revealed that the males scored again higher on pleasure before (M = 7.32) and
after the anxious movie (M = 5.05) than the females (at baseline M = 6.26, and after the anxious
movie M = 3.74). The arousal ratings were increased after the anxious mood induction (F(1,
34) = 76.38, p < .001, η
2
= .65). The mood induction and ambience group interaction was also
significant (F(1, 34) = 4.97, p = .032, η
2
= .13). As can be seen in Fig 4 (right figure), the arousal
scores of the elderly in the cozy ambience group increase more after the anxious movie than
the arousal scores of the elderly in the neutral ambience group. The effect of ambience group
on arousal was also significant (F(1, 34) = 8.45, p = .006, η
2
= .20). The elderly in the cozy ambi-
ence group scored higher on arousal at baseline and after the mood induction (as can be seen
from Fig 4). For the physiological measures, significant effects of the anxious mood induction
were found on SCR and HR. The SCR significantly increased from baseline (M = 4.45) to after
the mood induction (M = 4.78) in line with our expectations and the SAM ratings (F(1, 33) =
5.81, p = .022, η
2
= .15). For HR we found a significant gender x ambience group interaction
effect (F(1, 29) = 10.35, p = .003, η
2
= .26); further analysis revealed that males in the neutral
ambience group had a higher HR (M = 70.8) than the females (M = 62.3), whereas a reversed
effect was found for the cozy ambience group (males M = 62.0, and females M = 75.6).
Influence of lighting on mood
To analyze the influence that the activating and cozy ambiences had on mood (with respect to
that of the neutral ambience), we monitored the evolution of the affective measures, starting
from the moment the lighting was changed from the movie lighting setting to the cozy, activat-
ing or neutral setting (see Table 2). In particular, we resorted to what we call “Affective change
scores”. Affective change scores were calculated for every moment after the mood induction
Fig 4. Average pleasure and arousal scores before and after the sad movie (left) and anxious movie (right). The different bars represent the different
ambience groups. The error bars reflect the 95% CI.
doi:10.1371/journal.pone.0132732.g004
Lighting to Make You Feel Better
PLOS ONE | DOI:10.1371/journal.pone.0132732 July 20, 2015 12 / 22
and for every participant by subtracting the affective rating measured at the end of the mood
induction procedure from the affective rating measured after each of the five waiting periods
(i.e., ‘2 min’, ‘4 min’, ‘6 min’, ‘8 min’, and ‘10 min’). For the physiological measures, affective
change values were calculated by subtracting the median value of the signal over the last 120
seconds of the mood induction procedure from the median value measured during the 120 sec-
onds of each of the five waiting periods (i.e., ‘2 min’, ‘4 min’ , ‘6 min’, ‘8 min’, and ‘10 min’).
A general linear model was tested for each dependent variable (i.e., the affective change
scores for SAM and the affective change values for the physiological measures). Gender and
ambience group were included as between-subjects factors and ‘time’ (as discretized in the five
waiting periods described above) was included as a within-subjects factor. Since the previous
analysis (on the effectiveness of mood induction) revealed that the ambience groups were
sometimes influenced in a different way by the mood induction, we also calculated for each
participant and for each dependent variable the reaction to the movie. The latter was then
included as covariate in the analyses. The reaction to the movie was calculated by subtracting
the rating at baseline from the rating after mood induction for the affective scores, and by sub-
tracting the median values of the last 120 seconds of the baseline from the median value of the
last 120 second s of the mood induction for the physiological measures. The covariate was sig-
nificant in all analyses described below.
Influence of activating ambience on sad mood
Fig 5 presents the subjective pleasure and arousal change scores for participants exposed to the
neutral and activating ambiences after the sad mood induction. The pleasure change scores are
slightly higher for the participants in the activating ambience than for the participants in the
neutral ambience; this effect is, however, not significant. The arousal change scores are reduced
more under the neutral lighting than under the activating lighting; however also this effect is
not significant. No significant effect of gender or time on the affective change scores, nor a sig-
nificant interaction were found.
Fig 5. Average change scores with the error bars reflecting the 95% CI for pleasure (left) and arousal (right) after the sad mood induction. The
different bars represent the ambience to which the participants were exposed.
doi:10.1371/journal.pone.0132732.g005
Lighting to Make You Feel Better
PLOS ONE | DOI:10.1371/journal.pone.0132732 July 20, 2015 13 / 22
With respect to the physiological measures (shown in Fig 6) we found significant differences
between the activating and neutral ambience for both the SCR and HR measures. In line with
our expectations, the SCR of the participants immersed in the activating ambience increased
more compared to the SCR of the participants immersed in the neutral ambience (F(1, 31) =
4.30, p = .046, η
2
= .12). Also the effect of time reached significance (F(2.4, 75.7) = 6.21 p =
.002, η
2
= .17); the SCR increased during the 10 minutes after mood induction for both the par-
ticipants in the activating ambience and in the neutral ambience (see also Fig 6). The effect of
ambience on HR was also significant (F(1, 30) = 5.72, p = .023, η
2
= .16). The heart rate deceler-
ated for the participants in the neutral ambience, however not for the participants in the acti-
vating ambience (see again Fig 6), as such indicating a higher physiological arousal in the
activating ambience. No significant effects were found on the heart rate variability
measurements.
Influence of cozy ambience on anxious mood
The subjective pleasure and arousal change scores for all measurements after the anxious
mood induction are presented in Fig 7, for both the cozy and the neutral ambience group. The
pleasure change scores are higher for participants exposed to the cozy lighting as compared to
participants immersed in a neutral lighting; indeed, we found a significant effect of ambience
on pleasure (F(1, 33) = 5.68, p = .023, η
2
= .15). The arousal scores were significantly more
reduced under cozy lighting than under neutral lighting (F(1, 33) = 6.10, p = .019, η
2
= .16).
The effect of ambience group x gender on arousal was also significant (F(1, 33) = 4.37, p = .044,
η
2
= .12); the difference in arousal change score between the neutral and cozy ambience was
larger for the female participants than for the male partici pants (as is illustrated in Fig 8).
With respect to the physiological measures, we found a significant effect of ambience on
SCR (F(1, 32) = 4.43, p = .043 η
2
= .12). As expected and in line with the SAM arousal ratings,
the SCR affective change values of the participants in the cozy ambience group were signifi-
cantly lower than the SCR affective change values of the participants in the neutral ambience
Fig 6. Average change scores with the error bars reflecting the 95% CI for SCR (left) and HR (right) after the sad mood induction. The different bars
represent the ambience to which participants were exposed.
doi:10.1371/journal.pone.0132732.g006
Lighting to Make You Feel Better
PLOS ONE | DOI:10.1371/journal.pone.0132732 July 20, 2015 14 / 22
group (as is illustrated in Fig 9). The HR also decelerated after the anxious mood induction for
the participants in th e cozy ambience, while that was not the case for the participants in the
neutral ambience group (see also Fig 9). This effect was however not significant. No significant
effects were found on the HRV measures.
Discussion
In this study, we investigated whether atmospheres created with lighting (that is, ambiences
with a clearly recognized affective connotation) had the potential to improve a negative mood
of elderly. Specifically, we studied whether these affective ambiences were more effective in
countering negative moods than neutral ambiences, in terms of both increasing the elderly's
mood valence and stabilizing the arousal towards a more neutral level. According to both self-
reported and physiological measures, the low arousing, pleasant ambience (i.e., cozy) was
found to be more effective in reducing a high arousing negative mood state (i.e., anxiety) than
the neutral ambience. The high arousing, pleasant ambience (i.e., activating) was found to
increase the physiological arousal of sad elderly more than the neutral ambience, but this find-
ing was not supported by changes in SAM arousal ratings.
The effectiveness of the ambiences in countering negative moods was tested under controlled
conditions in a dedicated laboratory environment. The test participants (healthy elderly people)
were on average in a pleasant (M = 6.8 on a scale from 1 to 9) low arousing (M = 3.0 on a scale
from 1 to 9) mood state. We induced two negative mood states with a sad and anxious movie
excerpt. Both the anxious and the sad movie were successful in inducing negative valence; how-
ever, both movies also induced arousal. The anxious movie induced a high arousing negative
mood state as intended and in line with previous research [52], as verified through the analysis
of SAM arousal ratings and as supported by an increased electrodermal response. The sad
mood induction was intended to induce a low arousing negative mood state. In contrast with
previous research [52], we found that the movie excerpt from ‘The champ’ also induced arousal.
A possible explanation for this is that previous results were derived from observing mood
Fig 7. Average change scores with the error bars reflecting the 95% CI for pleasure (left) and arousal (right) after the anxious mood induction. The
different bars represent the different ambiences in which participants were immersed.
doi:10.1371/journal.pone.0132732.g007
Lighting to Make You Feel Better
PLOS ONE | DOI:10.1371/journal.pone.0132732 July 20, 2015 15 / 22
induction on younger adults. Our reference group instead consisted of elderly, for which the
response to mood induction may be different. Literature suggests that elderly may experience
more arousal when the sad mood induction is very relevant to their age group [62,63]. When
watching sad movies related to family loss, older people’s physiological arousal indeed increased
more than that of younger people [62,63]. In addition, also other affective responses may be
induced on top of sadness; Kunzmann and Gruhn [62] found that a family loss movie also
induced higher arousing mood states, as anxiety and anger, especially in older participants.
Despite the random assignment of the participants, while counterbalancing gender, we
found group differences in the reaction to the movies. Especially the elderly assigned to the
cozy ambience group had a more pronounced reaction to the anxious movie compared to the
elderly assigned to the neutral ambience group (note that both groups viewed the movie in the
same lighting condition). The more pronounced reaction might explain the differences we
found after the mood induction between the cozy and neutral ambience. To control for this
effect, we calculated the reaction to the mood induction for each participant separately (and
also for each dependent measure separately), and included it as covariate in the analysis. Part
of the differences found between the cozy and neutral ambiences were indeed explained by the
Fig 8. Average arousal change scores (also averaged over time), showing also the 95% CI as error
bars, after the anxious mood induction for the male and female participants separately. The different
bars represent the different ambiences in which participants were immersed.
doi:10.1371/journal.pone.0132732.g008
Lighting to Make You Feel Better
PLOS ONE | DOI:10.1371/journal.pone.0132732 July 20, 2015 16 / 22
covariate. However we still found that the elderly in the cozy ambience reported significantly
higher pleasure ratings and significantly lower arousal ratings than the elderly in the neutral
ambience. The SCR of the participants was also significantly reduced in the cozy ambience
compared to th e neutral ambience, even when the covariate was included in the analysis. The
effect of HR however was almost completely explained by the covariate, underpinning the
importance of including the covariate in the analyses.
We found a main effect of gender on the self-reported pleasure and arousal ratings after
mood induction and an interaction between ambience and gender on HR. These effects could be
completely explained by baseline differences, and so, the female and male participants did not
react differently to the affective movies (i.e., no significant mood induction x gender interactions
were found). Some studies suggest that female participants are more susceptible to mood influ-
ences than male participants [64,65]. These studies are however performed with young adults;
the few studies which investigated the effect of mood induction on elderly participants did not
find gender effects [66–68]. We however also found a significant effect of gender on the reaction
to the cozy versus the neutral ambience. Although the cozy ambience reduced the arousal scores
of both the female and male participants as compared to the neutral ambience, this reduction
was far more effective for the female participants. Knez and colleagues even found that lighting
can have reversed effects on the mood of females compared to males [39,40].
The activating ambience was found to be physiologically arousing compared to the neutral
ambience. No significant effects were found, though, on the self-reported levels of arousal
between the neutral and activating ambience; both participants immersed in the activating
ambience and participants exposed to the neutral ambience reported a similar reduction in
arousal during the waiting periods after the mood induction. The lack of effectiveness of the
activating ambience on the self-reported arousal may be due to a number of reasons. First, the
setup of the experiment might have worked against the psychological effect of the activating
ambience. In our protocol, participants went through five waiting periods of two minutes each,
during which they did nothing while sitting on a couch. On top of this, they were asked to limit
Fig 9. Average change scores, including the 95% CI as error bars, for SCR (a) and HR (b) after the anxious mood induction. The different bars
represent the different ambiences in which participants were immersed.
doi:10.1371/journal.pone.0132732.g009
Lighting to Make You Feel Better
PLOS ONE | DOI:10.1371/journal.pone.0132732 July 20, 2015 17 / 22
their movements not to hamper the recording of physiological measurements. Thus, it may be
the case that participants experienced the activating effect of the ambience, however, the exper-
imental task prevailed a more active stance and thus a more active feeling. Second, it may be
the case that the period of ten minutes was too short to realize a positive activating mood state.
Knez and colleagues [39–41] argue that longer exposure times of several hours, are needed to
generate consistent effects of indoor lighting on mood. Although the results of the cozy ambi-
ence suggest that consistent effects can be reached in short periods, it might require longer
exposure times to positively energize people.
We focused on the psychological effects of lighting on mood, however our ambiences did
not violate the known biological effects of lighting. The low arousing cozy ambience had light-
ing with a low illuminance and with a higher spectral density at the higher wavelengths (result-
ing from a lower CT of the functional white lighting and orange colored accent lighting). The
high arousing activating ambience had a higher illuminance with a higher spectral density at
shorter wavelengths (resulting from with a higher CT of the functional white lighting and cyan
colored accent lighting). So, if the lighting would have had any biological effect on mood, it
would have been in accordance to the psychological effect. The illuminance levels, however,
even in the activating ambience, are generally too low to reach a direct biological arousing
effect during day time.
We did not find any significant effect of atmosphere on the HRV measures. A possible
explanation may be found in the timeframes of 120 seconds for the analysis of the HRV data,
which may be too short. Periods of 5 minutes are recommended for short-term HRV record-
ings [60]. Such a longer recording time allows a more accurate calculation of the low frequency
component of HRV in the frequency domain (HRV-LF), which allows the calculation of nor-
mative values of HRV-HF and the HRV-HF/HRV-LF ratio. The setup of the experiment, how-
ever, did not allow the selection of longer timeframes for the analysis of the physiological data
because we wanted to measure the evolution of the mood over short time periods , to avoid lon-
ger waiting times possibly inducing boredom or frustration in the participants.
This investigation also has limitations, which should be acknowledged. First, we tested the
effect of the ambiences on induced negative moods and not on genuine negative moods.
Although research revealed that induced moods have similar properties and similar effects on
cognition and behavior as real negative moods [68], we know that the induced moods diminish
quickly after the mood induction. The effectiveness of the ambiences should be further tested
with elderly which are actually experiencing negative moods. Secondly, we only tested short-
term effects of the ambiences, while long-term effects are important to be investigated in order
to be able to claim that these ambien ces are fully beneficial for elderly. The effects of longer and
repeated exposure to ambiences still need to be determined. Thirdly, we used the Self-Assess-
ment Manikin (SAM) for measuring the self-perceived mood state of the elderly. This method
is quick and relatively low on cognitive effort, however at the cost of more detailed mood infor-
mation. As a result, we have no information on whether the elderly for instance experienced
mixed affective states after the mood induction, as might be the case after the sad mood induc-
tion. It is also unknown at this stage whether the setup of the experiment (with the five waiting
periods) induced negative affective states like boredom or frustration.
Another limitation of the current study is that we considered two ambiences which were on
average found to be cozy or activating by our target group. However, within our group of par-
ticipants large variations in intensity sensitivity and color perception may exist, due to the age-
related deterioration of the visual system [69]. Individual preferences and gender differences
may also have an impact on the optimality of ‘cozy’ and ‘activating’ ambiences. The usage of an
ambience considered ‘cosy’ or ‘activating’ by an average elderly may therefore result in subopti-
mal effectiveness of the ambience.
Lighting to Make You Feel Better
PLOS ONE | DOI:10.1371/journal.pone.0132732 July 20, 2015 18 / 22
Finally, we did not control for the contribution of separate lighting characteristics to the
measured mood effect, since the ambiences differed simultaneously in illumination, color tem-
perature and the colors used for the accent lighting. In this study, though, we were interested in
investigating if ambiences with a clear affective meaning could change the mood of people in
the direction of the affective meaning of the ambience. Next steps will be dedicated to link
lighting characteristics to their effect on mood, starting from the knowledge existing on the
relationship between lighting characteristics and atmosphere perception.
Conclusion
In this study, we showed that we can improve the mood of elderly by exposing them to lighting
ambiences with a clear, positive affective meaning. Sad elderly which were immersed in a posi-
tive high arousing (i.e., activating) ambience were physiologically more aroused than sad
elderly which were immersed in a neutral ambience. The anxious elderly could be effectively
calmed with a cozy ambience. Indeed, anxious elderly immersed in a cozy ambience were
physiologically less aroused and reported lower arousal scores and higher pleasure scores as
compared to anxious elderly immersed in a neutral ambience. First assessing how elderly per-
ceive the atmosphere of the ambiences [12] was found to be an efficient and reliable method to
find the lighting settings, which could effectively induce the desired mood change. To fully vali-
date our approach, the effectiveness of the ambience in improving mood should be further
tested with longer exposure times and with elderly with genuine negative moods. Once their
effectiveness is proven, we expect these ambiences to be beneficial also in contexts different
from the care center one, such as in hospital rooms.
Acknowledgments
We are grateful to the Vitalis WoonZorg Groep for assisting us in our research.
Author Contributions
Conceived and designed the experiments: AK BdR JR IH. Performed the experiments: AK.
Analyzed the data: AK. Contributed reagents/materials/analysis tools: AK. Wrote the paper:
AK BdR JR IH.
References
1. Aarts E, de Ruyter B. New research perspectives on Ambient Intelligence. J Ambient Intell Smart Envi-
ron. 2009 Jan 1; 1(1):5–14. doi: 10.3233/AIS-2009-0001
2. Dewsbury G, Taylor B, Edge M. Designing safe smart home systems for vulnerable people. In: Proctor
R, Rouncefield M, editors. Dependability in Healthcare Informatics. Edinburgh: Proceedings of the
First Dependability IRC Workshop, Edinburgh March;
3. Nay R. Nursing home residents’ perceptions of relocation. J Clin Nurs. 1995; 4(5):319–325. doi: 10.
1111/j.1365-2702.1995.tb00030.x
4. Lee DTF, Woo J, Mackenzie AE. A review of older people’s experiences with residential care place-
ment. J Adv Nurs. 2002; 37(1):19–27. doi: 10.1046/j.1365-2648.2002.02060.x
5. Fredrickson BL. Cultivating positive emotions to optimize health and well-being. Prev Treat. 2000; 3(1):
Article 1. doi: 10.1037/1522-3736.3.1.31a
6. Lyubomirsky S, King L, Diener E. The benefits of frequent positive affect: does happiness lead to suc-
cess? Psychol Bull. 2005; 131(6):803–855. doi: 10.1037/0033-2909.131.6.803
7. LeGates TA, Fernandez DC, Hattar S. Light as a central modulator of circadian rhythms, sleep and
affect. Nat Rev Neurosci. 2014 Jul; 15(7):443–454. doi: 10.1038/nrn3743
8. Küller R, Ballal S, Laike T, Mikellides B, Tonello G. The impact of light and colour on psychological
mood: a cross-cultural study of indoor work environments. Ergonomics. 2006 Nov 15; 49(14):1496–
1507. doi: 10.1080/00140130600858142
Lighting to Make You Feel Better
PLOS ONE | DOI:10.1371/journal.pone.0132732 July 20, 2015 19 / 22
9. Küller R, Mikellides B, Janssens J. Color, arousal, and performance—A comparison of three experi-
ments. Color Res Appl. 2009; 34(2):141–152. doi: 10.1002/col.20476
10. Kaiser PK. Physiological response to color: A critical review. Color Res Appl. 1984; 9(1):29–36. doi: 10.
1002/col.5080090106
11. Vogels IMLC. Atmoshere metrics. In: Westerink JHDM, Ouwerkerk M, Overbeek TJM, Pasveer WF, de
Ruyter B, editors. Probing experiece: From assessment of user emotions and behaviour to develop-
ment of products. The Netherlands: Springer; 2008. pp. 25–41.
12. Kuijsters A, Redi J, de Ruyter B, Seuntiëns P, Heynderickx I. Affective ambiences created with lighting
for older people. Light Res Technol. 2014; doi: 10.1177/1477153514560423
13. Custers P, de Kort Y, IJsselsteijn W, de Kruiff M. Lighting in retail environments: Atmosphere perception
in the real world. Light Res Technol. 2010 Sep 10; 42(3):331–343. doi: 10.1177/1477153510377836
14. Russell JA. Core affect and the psychological construction of emotion. Psychol Rev. 2003; 110(1):145–
172. doi: 10.1037/0033-295X.110.1.145
15. Cajochen C. Alerting effects of light. Sleep Med Rev. 2007 Dec; 11(6):453–64. doi: 10.1016/j.smrv.
2007.07.009
16. Wirz-Justice A, Benedetti F, Berger M, Lam RW, Martiny K, Terman M, et al. Chronotherapeutics (light
and wake therapy) in affective disorders. Psychol Med A J Res Psychiatry Allied Sci. 2005; 35(7):939–
944. doi: 10.1017/s003329170500437x
17. Mclntyre IM, Norman TR., Burrows GD., Armstrong SM. Human melatonin suppression by light is inten-
sity dependent. J Pineal Res. 1989; 6:149–156. doi: 10.1111/j.1600-079X.1989.tb00412.x
18. Lewy AJ., Sack RL. The dim light melatonin onset as a marker for circadian phase position. Chronobiol
Int. 1989; 6(1):93–102. doi: 10.3109/07420528909059144
19. Boyce P, Barriball E. Circadian rhythms and depression. Aust Fam Physician. 2010; 39(5):307–310.
20. Bedrosian TA, Nelson RJ. Influence of the modern light environment on mood. Mol Psychiatry. 2013
Jul; 18(7):751–757. doi: 10.1038/mp.2013.70
21. Shikder S, Mourshed M, Price A. Therapeutic lighting design for the elderly: a review. Perspect Public
Health. 2012 Nov; 132(6):282–291. doi: 10.1177/1757913911422288
22. Golden RN, Gaynes BN, Ekstrom RD, Hamer RM, Jacobsen FM, Suppes T, et al. The efficacy of light
therapy in the treatment of mood Ddsorders: A Review and Meta-Analysis of the Evidence. Am J Psy-
chiatry. 2005; 162(4):656–662. doi: 10.1176/appi.ajp.162.4.656
23. Warman VL, Dijk D-J, Warman GR, Arendt J, Skene DJ. Phase advancing human circadian rhythms
with short wavelength light. Neurosci Lett. 2003; 342(1–2):37–40. doi: 10.1016/s0304-3940(03)00223-
4
24. Thapan K, Arendt J, Skene DJ. An action spectrum for melatonin suppression: evidence for a novel
non-rod, non-cone photoreceptor system in humans. J Physiol. 2001; 535(1):261–267. doi: 10.1111/j.
1469-7793.2001.t01-1-00261.x
25. Jung CM, Khalsa SBS, Scheer FAJL, Cajochen C, Lockley SW, Czeisler CA, et al. Acute effects of
bright light exposure on cortisol levels. J Biol Rhythms. 2010 Jun 1; 25(3):208–216. doi: 10.1177/
0748730410368413
26. Yoto A, Katsuura T, Iwanaga K, Shimomura Y. Effects of object color stimuli on human brain activities
in perception and attention referred to EEG alpha band response. J Physiol Anthropol. 2007; 26
(3):373–379. doi: 10.2114/jpa2.26.373
27. Ali MR. Pattern of EEG recovery under photic stimulation by light of different colors. Electroencephalogr
Clin Neurophysiol. 1972; 33(3):332–335. doi: 10.1016/0013-4694(72)90162-9
28. Wilson GD. Arousal properties of red versus green. Percept Mot Skills. 1966; 23(3):947–949. doi: 10.
2466/pms.1966.23.3.947
29. Jacobs KW, Hustmeyer FE. Effects of four psychological primary colors on GSR, heart rate and respira-
tion rate. Percept Mot Ski. 1974; 38:763–766. doi: 10.2466/pms.1974.38.3.763
30. Gerard RM. Differential Effects of Coloured Lights on Psycho-Physiological Functions. 1958;Ph.D.
Diss.
31. Valdez P, Mehrabian A. Effects of color on emotions. J Exp Psychol Gen. 1994; 123(4):394–409. doi:
10.1037/0096-3445.123.4.394
32. Nourse JC, Welch RB. Emotional attributes of colour: a comparison of violet and green. Percept Mot
Skills. 1971; 32(2):403–406. doi: 10.2466/pms.1971.32.2.403
33. Bellizzi JA, Hite RE. Environmental color, consumer feelings, and purchase likelihood. Psychol Mark.
1992; 9(5):347–363. doi: 10.1002/mar.4220090502
Lighting to Make You Feel Better
PLOS ONE | DOI:10.1371/journal.pone.0132732 July 20, 2015 20 / 22
34. Kwallek N, Lewis CM, Lin-Hsiao JWD, Woodson H. Effects of nine monochromatic office interior colors
on clerical tasks and worker mood. Color Res Appl. 1996; 21(6):448–458. doi: 10.1002/(sici)1520-6378
(199612)21:6<448::aid-col7>3.0.co;2-w
35. Stone J. N. Designing effective study environments. Environ Psychol. 2001; 21:179–190. doi: 10.1006/
jevp.2000.0193
36. Mikellides B. Color and physiological arousal. J Archit Plann Res. 1990; 7:13–20.
37. Baron RA, Rea MS, Daniels SG. Effects of indoor lighting (illuminance and spectral distribution) on the
performance of cognitive tasks and interpersonal behaviors: The potential mediating role of positive
affect. Motiv Emot. 1992; 16(1):1–33.
38. Knez I. Effects of colour of light on nonvisual psychological processes. J Environ Psychol. 2001 Jun; 21
(2):201–208. doi: 10.1006/jevp.2000.0198
39. Knez I. Effect of indoor lighting on mood and cognition. J Environ Psychol. 1995;( 15):39–51. doi: 10.
1016/0272-4944(95)90013-6
40. Knez I, Enmarker I. Effects of office lighting on mood and cognitive performance and a gender effect in
work-related judgment. Environ Behav. 1998; 30(4):553–567. doi: 10.1177/001391659803000408
41. Knez I, Kers C. Effects of indoor lighting, gender, and age on mood and cognitive performance. Environ
Behav. 2000 Nov 1; 32(6):817–831. doi: 10.1177/0013916500326005
42. Veitch J, Newsham G, Boyce P, Jones C. Lighting appraisal, well-being and performance in open-plan
offices: A linked mechanisms approach. Light Res Technol. 2008 Jun 1; 40(2):133–151. doi: 10.1177/
1477153507086279
43. Liu XY, Luo MR, Li H. A study of atmosphere perceptions in a living room. Light Res Technol. 2014 Apr
1; doi: 10.1177/1477153514528934
44. Russell JA, Ward LM, Pratt G. Affective quality attributed to environments: a factor analytic study. Envi-
ron Behav. 1981 May 1; 13(3):259–288. doi: 10.1177/0013916581133001
45. Russell JA, Pratt G. A description of the affective quality attributed to environments. J Pers Soc Psy-
chol. 1980; 38(2):311–322. doi: 10.1037/0022-3514.38.2.311
46. Wang HH, Luo MR, Liu P, Yang Y, Zheng Z, Liu X. A study of atmosphere perception of dynamic col-
oured light.
Light Res Technol. 2013 Oct 21; doi: 10.1177/1477153513506591
47. Seuntiëns PJH, Vogels IMLC. Atmosphere creation: the relation between atmosphere and light charac-
teristics. In: The 6th International Conference on Design and Emotion. 2008.
48. Huldtgren A, Katsimerou C, Kuijsters A, Redi JA, Heynderickx I. Design considerations for adaptive
lighting to improve seniors’ mood. In: International Conference On Smart homes and health Telematics.
2015.
49. Gerrards-Hesse A, Spies K, Hesse FW. Experimental inductions of emotional states and their effective-
ness: A review. Br J Psychol. 1994; 85(1):55–78. doi: 10.1111/j.2044-8295.1994.tb02508.x
50. Martin M. On the induction of mood. Clin Psychol Rev. 1990 Jan; 10(6):669–697. doi: 10.1016/0272-
7358(90)90075-L
51. Westermann R, Spies K, Stahl Gün, Hesse FW. Relative effectiveness and validity of mood induction
procedures: a meta-analysis. Eur J Soc Psychol. 1996; 26(4):557–580. doi: 10.1002/(sici)1099-0992
(199607)26:4<557::aid-ejsp769>3.0.co;2-4
52. Gross JJ, Levenson RW. Emotion elicitation using films. Cogn Emot. 1995; 9(1):87–108. doi: 10.1080/
02699939508408966
53. Zeffirelli F, Lovell D. The Champ. Culver City, CA: MGM/Pathe Home Video; 1979.
54. Demme J, Saxon E, Utt K, Bozman R. The silence of the lambs. New York, NY: Orion Pictures; 1991.
55. Bradley MM, Lang PJ. Measuring emotion: The self-assessment manikin and the semantic differential.
J Behav Ther Exp Psychiatry. 1994; 25(1):49–59. doi: 10.1016/0005-7916(94)90063-9
56. Gomez P, Zimmermann P, Guttormsen-Schär S, Danuser B. Valence lasts longer than arousal persis-
tence of induced moods as assessed by psychophysiological measures. J Psychophysiol. 2009; 23
(1):7–17. doi: 10.1027/0269-8803.23.1.7
57. Bradley MM, Codispoti M, Cuthbert BN, Lang PJ. Emotion and motivation I: Defensive and appetitive
reactions in picture processing. Emotion. 2001; 1(3):276–298. doi: 10.1037/1528-3542.1.3.276
58. Lang PJ, Greenwald MK, Bradley MM, Hamm AO. Looking at pictures: Affective, facial, visceral, and
behavioral reactions. Psychophysiology. 1993; 30(3):261–273. doi: 10.1111/j.1469-8986.1993.
tb03352.x
59. Gomez P, Stahel WA, Danuser B. Respiratory responses during affective picture viewing. Biol Psychol.
2004; 67(3):359–373. doi: 10.1016/j.biopsycho.2004.03.013
Lighting to Make You Feel Better
PLOS ONE | DOI:10.1371/journal.pone.0132732 July 20, 2015 21 / 22
60. Task Force of the European Society of Cardiology and North American Society of Pacing and Electro-
physiology. Heart rate variability: standards of measurement, physiological interpretation and clinical
use. Circulation. 1996; 93(5):1043–1065. Greater sadness reactivity in late lifeGreater sadness reactiv-
ity in late life
61. Perakakis P, Joffily M, Taylor M, Guerra P, Vila J. KARDIA: A Matlab software for the analysis of cardiac
interbeat intervals. Comput Methods Programs Biomed. 2010; 98(1):83–89.
62. Kunzmann U, Grühn D. Age differences in emotional reactivity: the sample case of sadness. Psychol
Aging. 2005; 20(1):47–59. doi: 10.1037/0882-7974.20.1.47
63. Seider BH, Shiota MN, Whalen P, Levenson RW. Greater sadness reactivity in late life. Soc Cogn
Affect Neurosci. 2010; doi: 10.1093/scan/nsq069
64. Albersnagel FA. Velten and musical mood induction procedures: A comparison with accessibility of
thought associations. Behav Res Ther. 1988; 26(1):79–95. doi: 10.1016/0005-7967(88)90035-6
65. Robins CJ. Development of experimental mood induction procedures for testing personality-event inter-
action models of depression. J Clin Psychol. 1988 Nov 1; 44(6):958–963. doi: 10.1002/1097-4679
(198811)44:6<958::AID-JCLP2270440618>3.0.CO;2-D
66. Chou K-L, Lee TMC, Ho AHY. Does mood state change risk taking tendency in older adults? Psychol
Aging. 2007; 22(2):310–318. doi: 10.1037/0882-7974.22.2.310
67. Feng MC, Courtney CG, Mather M, Dawson ME, Davison GC. Age-related affective modulation of the
startle eyeblink response: older adults startle most when viewing positive pictures. Psychol Aging.
2011 Sep; 26(3):752–60. doi: 10.1037/a0023110
68. Fox LS, Knight BG, Zelinski EM. Mood induction with older adults: A tool for investigating effects of
depressed mood. Psychol Aging. 1998; 13(3):519–523. doi: 10.1037/0882-7974.13.3.519
69. Johnson ML. The cambridge handbook of age and ageing. Cambridge Handbooks Psychol. 2005.
Lighting to Make You Feel Better
PLOS ONE | DOI:10.1371/journal.pone.0132732 July 20, 2015 22 / 22
