Food Quality and Preference 69 (2018) 1–9

Contents lists available at ScienceDirect

Food Quality and Preference

journal homepage: www.elsevier.com/locate/foodqual

The paradox of warmth: Ambient warm temperature decreases preference T

for savory foods
⁎
Kosuke Motokia,b, , Toshiki Saitoa, Rui Nouchia,c,d, Ryuta Kawashimaa,c, Motoaki Sugiuraa,d
a
Institute of Development, Aging and Cancer, Tohoku University, Sendai, Japan
b
Japan Society for the Promotion of Science, Tokyo, Japan
c
Smart-Aging Research Center, Sendai, Japan
d
International Research Institute of Disaster Science, Tohoku University, Sendai, Japan

A R T I C LE I N FO A B S T R A C T

Keywords: Grocery stores and restaurants can control ambient temperatures using air-conditioning. Although questions
Ambient temperature about how ambient temperature affects consumers’ evaluations of foods are relevant to many food-based
Warmth businesses, they remain largely unanswered, and there are contradictory hypotheses regarding the influence of
Food preference ambient temperature on food preferences. Embodiment theory suggests that ambient warm temperature in-
Perceived food temperature
creases food preferences, whereas thermoregulation theory suggests that ambient warm temperature may de-
Sensory marketing
crease food preferences to lower the body’s temperature by limiting food intake. However, neither of these
explanations considers food category. Given that food varies according to whether it is usually eaten warm or
cool, the thermoregulation theory leads to the hypothesis that people may try to regulate their body temperature
by preferring foods that should be eaten cool and avoiding those that should be eaten warm. To resolve these
contradictory hypotheses, this study investigated the effect of warm ambient temperature on preferences for
different categories of food. In total, 52 participants in a room with either warm or cool ambient temperature
reported preferences for four categories of food images: vegetables, fruits, sweets, and savory foods. The foods
were grouped into warm foods (i.e., foods perceived as having a warm temperature: savory foods), cold foods
(i.e., foods perceived as having a cool temperature: fruits), and control foods (vegetables and sweets). The results
indicated that ambient warm temperatures decreased preferences for savory foods but did not affect preferences
for the other foods. The decreased preference for savory foods in warm ambient temperature was based on
perceived food temperature but not on tastiness or healthfulness. These findings are the first to establish the
effect of food temperature on food preference in warm ambient conditions. Incorporation of food temperature
into thermoregulation theory can advance understanding of the sensory influences on consumer behaviors.

1. Introduction 1.1. Sensory influences on consumer behaviors

Consumers view a variety of food images in grocery stores (e.g.,
food advertising, food packaging) and, especially in Asian cultures, are
exposed to such images on the menus of restaurants as they consider
what to order. Although the ambient temperature of the store may in-
fluence consumer behavior and can be controlled by store managers,
neither its effects on food-group preferences nor its underlying me-
chanisms are understood. Thus, an understanding of the effects of
ambient temperature may improve our theoretical understanding of,
and practical knowledge about, sensory influences on food preferences.
There is little evidence regarding how ambient temperatures influ-
ence consumer evaluations of foods. Sensory input (visual, auditory,
olfactory, tactile, and gustatory) plays a crucial role in food-related
consumer behavior (Krishna, 2012; Krishna & Schwarz, 2014; Spence,
2012). Store design, including electric lighting, background music, food
smell, and ambient temperature, influences consumers’ evaluations of
food (Krishna, Morrin, & Sayin, 2013; North, Hargreaves, &
McKendrick, 1999; Zwebner, Lee, & Goldenberg, 2013). For example,
in-store national music leads to increased selection of the associated
wine (e.g., German music and German wine selection; (North et al.,
1999). Scents in printed food advertisements increase salivation, desire
to eat, and food intake (Krishna et al., 2013). Despite research on the

⁎
Corresponding author at: Institute of Development, Aging and Cancer, Tohoku University, Seiryo-machi 4-1, Aoba-ku, Sendai 980-8575, Japan.
E-mail address: kosuke.motoki.p2@dc.tohoku.ac.jp (K. Motoki).

https://doi.org/10.1016/j.foodqual.2018.04.006
Received 11 December 2017; Received in revised form 18 April 2018; Accepted 18 April 2018
Available online 25 April 2018
0950-3293/ © 2018 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/BY-NC-ND/4.0/).
K. Motoki et al.
effects of sensory input on food-related consumer behavior, the effects
of ambient temperature on consumer evaluations of foods are not well
understood.
1.2. Ambient warm temperature may increase food preference
Although the effect of ambient temperature on food preferences has
not been extensively investigated, the relevant theory indicates that warm
ambient temperature may increase such preferences. According to embo-
diment theory, abstract psychological concepts (e.g., emotional warmth,
positive reactions) are metaphorically grounded in concrete physical ex-
periences (e.g., physical warmth) (Barsalou, 2010). This metaphorical
logic suggests that warmth triggers positive reactions, which then induce
congruent positive evaluations. In fact, physical warmth can, indeed, en-
hance positive assessments of others (Williams & Bargh, 2008). Based on
this theory, a previous study demonstrated the existence of a temperature
premium (warm temperatures lead to more positive evaluations of pro-
ducts) (Zwebner et al., 2013). Exposure to warmth increased the perceived
value of products, including food (e.g., cupcakes) and beverages (e.g.,
milk, coffee) (Zwebner et al., 2013). These results indicate that ambient
warm temperature may increase food preferences.
1.3. Ambient warm temperature may decrease food preferences
Thermoregulation theory suggests that ambient warm temperature
decreases food preference. According to this theory, body temperature is
regulated and homeostasis is maintained by adjusting food intake in re-
sponse to temperature change (Terrien, Perret, & Aujard, 2011). The basic
metabolic rate is lower in prolonged warm temperature, and this may lead
to reduced appetite and food intake to lower the body’s temperature
(Terrien et al., 2011). Rats consume less during periods of prolonged warm
temperature than during periods of prolonged cool temperature (Brobeck,
1948). People who live in warm climates consume less energy from food
(Johnson & Kark, 1947). Although the volume of food intake is propor-
tionate to the degree to which an individual prefers that food, higher
(lower) levels of food preference before eating lead to more (less) food
intake (Sørensen, Møller, Flint, Martens, & Raben, 2003). Based on this
evidence, thermoregulation theory suggests that ambient warm tempera-
ture decreases food preference.
1.4. The influence of ambient warm temperature on food preferences is
dependent on food category
Thermoregulation theory may lead to an alternative hypothesis
when considering foods that are usually eaten warm versus those that
are usually eaten cool. Although food intake regulates body tempera-
ture, its effect may be influenced by the perceived temperature of the
food. In animals, the intake of cold food reduces body temperature
more than the intake of warm food (Nicol & Young, 1989). In humans,
social exclusion (i.e., social coolness) is associated with feeling cool,
leading to an increased preference for warm liquids (coffee, soup)
(Zhong & Leonardelli, 2008), perhaps in an attempt at thermoregula-
tion (IJzerman, 2015). Moreover, consumption of a cool drink reduced
negative states (aggression) induced by an uncomfortably warm am-
bient temperature (Baron & Bell, 1976). The findings of this study also
implied that people prefer cold food in a warm ambient environment.
Thus, in warm temperature, humans are likely to devalue warm food
and prefer cold food to efficiently regulate their body temperature.
Although the categorization of food groups varies across studies, food
categories generally consist of four groups: vegetables, fruits, sweets,
and savory foods (Blechert, Meule, Busch, & Ohla, 2014; Epstein et al.,
2015; Howard, Gottfried, Tobler, & Kahnt, 2015; Padulo et al., 2017).
Fruits are generally perceived as cool, apart from a few such, as chili
peppers, whereas savory foods (non-sweet and salty/umami) (Howard
et al., 2015; Keast & Breslin, 2003; Olsen, Ritz, Hartvig, & Møller,
2011), such as pizza and fried potatoes, are generally perceived as
2
Food Quality and Preference 69 (2018) 1–9
warm. Fruit is often served cool, and savory foods are often cooked by
heat (e.g., baking, boiling, and frying). If fruits are perceived as cool
and savory foods are perceived as warm, ambient warm temperature
should increase the preference for fruit and decrease the preference for
savory foods.
1.5. The present study
To resolve the contradictory hypotheses about the effects of warm
temperature on food preferences, this study investigated the effect of
warm ambient temperature on preferences for different categories of
food images and examined the factors that influence preference shifts.
In total, 52 participants in rooms with a relatively warm or cool am-
bient temperature reported their preferences for four categories of food
images: vegetables, fruits, sweets, and savory foods. These categories
were classified as warm (savory foods: those perceived as having a
warm temperature), cool (fruits: those perceived as having a cool
temperature), and control groups within the overarching categories of
hedonic/sinful (sweet) and utilitarian/virtuous (vegetables) groups. To
identify the underlying mediators of preference shifts, participants also
evaluated perceived food temperature, a candidate mediator, and the
two control factors that potentially influence food preferences (tastiness
and healthfulness) for each food.
Based on the different theories, we developed three hypotheses re-
garding the effects of ambient temperature on preferences for different
categories of food images. Based on embodiment theory, ambient warm
temperature increases food preferences regardless of category (H1).
Based on thermoregulation theory, ambient warm temperature de-
creases food preferences regardless of food category (H2a). The in-
corporation of food temperature into thermoregulation theory suggests
that ambient warm temperature may influence food preferences as a
function of the temperature of the food category in question. If fruits
were perceived as cool and savory foods were perceived as warm,
ambient warm temperature would be expected to increase preferences
for fruits and decrease those for savory foods (H2b).
• H1. Ambient warm (vs. cool) temperature increases food pre-
ferences regardless of food category (vegetables, fruits, sweets, and
savory foods).
• H2a. Ambient warm (vs. cool) temperature decreases food pre-
ferences regardless of food category (vegetables, fruits, sweets, and
savory foods).
• H2b. When fruits are perceived as cool and savory foods are per-
ceived as warm, ambient warm temperature will increase pre-
ferences for fruits, which are perceived as having a cool tempera-
ture, and decrease those for savory foods, which are perceived as
having a warm temperature.
2. Methods
2.1. Participants
In total, 52 healthy university students (31 females,
Mage = 21.02 ± 1.90) were recruited using a bulletin board notice and
a student mailing list. This study was approved by the Ethics Committee
of the School of Medicine at Tohoku University and was conducted in
accordance with the Declaration of Helsinki.
2.2. Design
The study had a 2 (temperature: warm/cool) × 4 (food groups:
vegetables, fruits, sweets, and savory foods) mixed-subjects design,
with temperature as the between-subjects factor and food groups as the
within-subject factor. The primary outcome was food preferences for
each category of food images.
K. Motoki et al. Food Quality and Preference 69 (2018) 1–9

2.3. Ambient warm vs. cool temperature

Groups of between two and six participants were randomly allo-

cated to a relatively warm or cool room that could accommodate a
maximum of 10 people. The ambient warm temperature was set at
27–30 °C (MWarm = 28.90 ± 1.39), and the ambient cool temperature
was set at 20–23 °C (MCool = 21.81 ± 0.75). Both temperature condi-
tions were slightly warm/cool and within the comfortable range to
establish ecological validity (i.e., the food industry seems to adjust
temperature to only a small degree). The temperature disparity (the
difference between the cool and warm conditions) was comparable to
that used in previous studies (Bell & Baron, 1977; Huang, Zhang, Hui, &
Wyer, 2014; IJzerman & Semin, 2009). The baseline temperature room,
where the ethical issues were explained and the basic data for each
participant were collected, was set at 24–27 °C.
A paired t-test was used to examine differences in the actual tempera-
tures (°C) of the warm and cool rooms. The temperature under the warm
condition was significantly higher than that under the cool condition
(MWarm = 28.90 °C vs. MCool = 21.81 °C, t50 = 22.603, p < 0.0001).
Differences in the perceived warmth of the warm and cool conditions were
analyzed with a paired t-test. Data for three participants were missing as a
result of not providing a response to the following question: “How warm do
you feel?” The perceived warmth, scored from 1 = cool to 7 = warm, was
greater under the warm condition than under the cool condition
(MWarm = 5.167 vs. MCool = 2.760, t47 = 9.379, p < 0.001). Additionally, Fig. 1. Examples of food images in each category.
we performed a one-sample t test to determine whether the subjective
perception of warmth significantly differed from neutral (a rating of 4 on a
7-point Likert scale). Participants in the warm room felt warm
(t23 = 5.935, p < 0.001), whereas those in the cool room felt cool
(t24 = −7.464, p < 0.001). The question “How do you feel about the
current temperature?” was used to evaluate attitudes toward temperature
and was scored from 1 = negative to 7 = positive. Attitudes did not differ
between conditions (MWarm = 4.083 vs. MCool = 3.520, t47 = 1.408,
p = 0.166). General emotional state was evaluated with the question
“What is your emotional state?” and scored from 1 = negative to 7 = po-
sitive; there were no significant differences between conditions in this re-
gard (MWarm = 4.042 vs. MCool = 4.120, t47 = −0.236, p = 0.815). These
data indicate that the temperature manipulation was successfully im-
plemented.

2.4. Food evaluation task

Forty food images were selected from food-pics (Blechert et al.,

2014) and categorized into four categories: vegetables, fruits, sweets,
and savory foods (Fig. 1). These categories were classified as warm
(savory foods), cool (fruits), and control (vegetables and sweets). Each
food category included 10 images. Participants rated the 40 different
foods in terms of tastiness (“How tasty is this food?”), healthfulness
(“How healthy is this food?”), and overall preference (“How much,
overall, do you like this food?”). These paradigms are similar to those
used in previous studies (Hare, Camerer, & Rangel, 2009; Motoki, Saito, Fig. 2. Participants in a warm or cool room rated foods at their own pace using
Nouchi, Kawashima, & Sugiura, 2018). The rating procedure used a a seven-point Likert scale across four separate blocks (tastiness, healthfulness,
seven-point Likert scale ranging from 1 (not at all) to 7 (very much) for overall liking, food temperature). Food blocks and items were randomized
each food item. Participants also rated perceived food temperature across subjects. Foods were categorized as fruits, sweets, savory foods, and
vegetables.
(“How warm is this food?” from 1 = cool to 7 = warm) for each food
item. The rating procedure was self-paced, and a fixation cross ap-
peared during the inter-stimulus interval for 0.5 s. All rating procedures 3.2. Fruits were rated as cooler and savory foods were rated as warmer than
were conducted using PsychoPy (Peirce, 2007). An example of the task other food categories
is illustrated in Fig. 2. All participants were required to fast for at least
3 h prior to the experiment (Hare et al., 2009; Sullivan et al., 2015). To assess whether fruits and savory foods were rated as cooler or
warmer than foods in other categories, generalized linear mixed models
3. Analysis and results (GLMMs) were applied to the data. A GLMM allows the nesting of a
level 1 variable (participant) within a level 2 variable (food) and has
3.1. Basic statistics the advantage of efficiently utilizing all available data in a set. The
dependent variable in the GLMMs was perceived food temperature.
Table 1 shows the basic statistics of the subjective ratings. Food-category dummy variable were set as fixed factors, and

3
K. Motoki et al. Food Quality and Preference 69 (2018) 1–9

Table 1
Subjective ratings.
Temperature Rating Food category

Savory foods Vegetables Sweets Fruits

Warm Liking 4.83 (1.36) 4.34 (1.45) 3.37 (1.76) 5.44 (1.29)
Tastiness 5.11 (1.42) 4.04 (1.35) 4.13 (1.77) 5.27 (1.24)
Healthfulness 3.48 (1.59) 5.66 (1.14) 2.29 (1.14) 5.71 (0.92)
Food hotness 4.74 (1.47) 3.10 (1.02) 3.30 (1.07) 2.54 (0.80)

Cool Liking 5.21 (1.36) 4.12 (1.70) 3.60 (1.84) 5.30 (1.35)
Tastiness 5.30 (1.37) 3.96 (1.70) 4.20 (1.55) 4.93 (1.31)
Healthfulness 3.42 (1.62) 5.74 (1.19) 2.13 (1.01) 5.65 (1.07)
Food hotness 4.70 (1.56) 3.32 (1.17) 3.40 (1.16) 2.87 (0.97)

Note: Values are presented as mean (standard deviation).

participants were set as random factors. When the food-category
dummy variables were significant, the other GLMMs were performed as
post hoc analyses to test for the effects of each pair of food categories on
perceived food temperature treating each food category as a dummy
variable (e.g., savory = 1, fruits = 0) and as a fixed factor and treating
participants as random factors. All statistical analyses were conducted
using R software (R Core Development Team, Austria, Vienna). The
lme4 package in R was used for the GLMMs (Bates, Mächler, Bolder &
Walker, 2014).
The food category dummy variables were significant in that savory
foods, vegetables, and sweets were perceived as significantly warmer
than fruits (savory: B = 1.449, SE = 0.051, p < 0.001; sweets:
B = 0.465, SE = 0.051, p < 0.001; vegetables: B = 0.362, SE = 0.050,
p < 0.001). Post hoc analysis indicated that savory foods were per-
ceived as warmer than fruits (B = 1.261, SE = 0.048, p < 0.001),
sweets (B = 0.995, SE = 0.053, p < 0.001), and vegetables
(B = 0.917, SE = 0.055, p < 0.001). Fruits were perceived as cooler
than sweets (B = −0.611, SE = 0.056, p < 0.001) and vegetables
(B = −0.488, SE = 0.053, p < 0.001).
These results indicate that perceived food temperature differs by
food category; savory foods were perceived as warmer, and fruits were
perceived as cooler.
3.3. Effects of ambient temperature on food preferences
To assess the effects of ambient temperature on food evaluations,
GLMMs were applied to the data. The dependent variable in the GLMMs
was food preference. Temperature was treated as a dummy variable
(warm = 1, cool = 0) and set as the fixed factor, whereas participants
were set as random factors. To ensure completeness, additional GLMMs
were run treating other evaluations of the foods (tastiness, healthful-
ness, and perceived food temperature) as dependent variables.
Ambient temperature did not influence food preferences
(B = −0.036, SE = 0.083, p = 0.666), tastiness (B = 0.025,
SE = 0.093, p = 0.791), or healthfulness (B = 0.025, SE = 0.063,
p = 0.686) ratings. Warm ambient room temperature tended to in-
crease perceived food temperature (B = −0.110, SE = 0.061,
p = 0.073).
These results indicate that ambient temperature did not influence
overall food preferences. Any observed effects may have been depen-
dent on food category.
3.4. Effects of ambient temperature and food group on food preferences
To assess the effects of the interaction between ambient temperature
and food group on preferences, GLMMs were applied to the data. Food
preference was treated as the dependent variable in the GLMMs, and
the temperature dummy variable (warm = 1, cool = 0), food-category
dummy variable, and their interaction were set as fixed factors.
Participants were set as random factors. To ensure completeness,
additional GLMMs were run treating other evaluations of the foods
(tastiness, healthfulness, and perceived temperature) as the dependent
variables.
When the interaction term was significant, additional GLMMs were
used in a post hoc analysis to determine the effects of temperature on
each food evaluation, with the temperature treated as a dummy vari-
able (warm = 1, cool = 0) and set as the fixed factor and participant set
as the random factor. All statistical analyses were conducted using R
software (R Core Development Team, Austria, Vienna). The lme4
package in R was used for the GLMMs (Bates et al., 2014).
3.4.1. Effects of ambient temperature on food preferences are dependent on
food category
The interaction between ambient temperature and food group in-
fluenced food preferences. Participants in the ambient warm tempera-
ture room expressed significantly decreased preferences for sweets and
savory foods compared with fruits (warm × sweets: B = −0.212,
SE = 0.107, p = 0.048; warm × savory food: B = −0.212, SE = 0.107,
p = 0.004) but not compared with vegetables (warm × vegetables:
B = 0.052, SE = 0.107, p = 0.628).
To investigate these interactions in detail, post hoc analyses were
conducted to test the effects of temperature on preferences for savory
foods, sweets, and fruits separately. Ambient temperature did not in-
fluence preferences for sweets (B = −0.123, SE = 0.181, p = 0.495) or
fruits (B = 0.104, SE = 0.159, p = 0.514). Ambient warm temperature
significantly decreased preferences for savory foods (B = −0.278,
SE = 0.126, p = 0.028).
These results indicate that the influence of ambient temperature on
food preference varied with food category. Warm (vs. cool) ambient
temperature decreased preferences for savory foods but not for other
foods (Fig. 3).
3.4.2. Effects of ambient temperature on tastiness are dependent on food
category
The interaction between ambient temperature and food group in-
fluenced evaluations of tastiness. Participants in the ambient warm (vs.
cool) temperature room rated sweets and savory foods as significantly
less tasty (warm × sweets: B = −0.268, SE = 0.110, p = 0.015;
warm × savory food: B = −0.342, SE = 0.110, p = 0.002), but no ef-
fect was observed on the tastiness ratings of vegetables (B = −0.160,
SE = 0.110, p = 0.146).
To investigate these interactions in detail, a post hoc analysis was
conducted to test the effects of temperature on the tastiness of savory
foods, sweets, and fruits. Warm ambient temperature did not affect the
tastiness ratings of sweets (B = −0.047, SE = 0.17, p = 0.789) or
savory foods (B = −0.138, SE = 0.148. p = 0.350), but a minor in-
crease in the tastiness ratings for fruits was observed (B = 0.261,
SE = 0.153, p = 0.088).
These results indicate that the influence of ambient temperature on
tastiness varied with food category. Warm (vs. cool) ambient

4
K. Motoki et al.
Fig. 3. Decreased preferences for savory foods in warm temperatures. references for savory food were decreased in warm (vs. cool) temperatures, whereas pre-
ferences for the other food categories were not. Error bar represents standard deviation. *p < 0.05.
temperature slightly increased tastiness ratings for fruits but not for
other foods.
3.4.3. Effects of ambient temperature on healthfulness are dependent on
food category
The interaction between ambient temperature and food group did
not influence healthfulness ratings (warm × vegetables: B = −0.070,
SE = 0.075, p = 0.347; warm × sweets: B = 0.054, SE = 0.075,
p = 0.468; warm × savory food: B = 0.003, SE = 0.075, p = 0.965).
These results indicate that ambient temperature did not influence
the healthfulness ratings, regardless of food category.
3.4.4. Effects of ambient temperature on perceived food temperature are
dependent on food category
The interaction between ambient temperature and food group in-
fluenced perceived food temperature. Participants in the ambient warm
(vs. cool) temperature room reported significantly increased perceived
food temperatures for savory foods (warm × savory foods: B = 0.259,
SE = 0.102, p = 0.011) but not for other foods (warm × sweets:
B = 0.166, SE = 0.102, p = 0.105; warm × vegetables: B = 0.077,
SE = 0.102, p = 0.450).
To investigate these interactions in detail, a post hoc analysis was
conducted to test the effects of temperature on the perceived food
temperature of savory foods and fruits. Warm (vs. cool) ambient tem-
perature did not influence the perceived temperature of savory foods
(B = 0.022, SE = 0.088, p = 0.804) but showed a slight tendency to
increase the tastiness of fruits (B = −0.366, SE = 0.183, p = 0.046).
Similar results were obtained when age and sex were included as
covariates.
3.5. Does perceived food temperature influence preferences for savory foods
and tastiness ratings for fruits?
Ambient warm temperature decreased preferences for savory foods
and increased fruit tastiness ratings. According to our understanding of
physiological thermoregulation, cold food intake efficiently decreases
body temperature, and the perceived temperature of foods should
5
Food Quality and Preference 69 (2018) 1–9
contribute to an evaluation shift. An additional GLMM was run to in-
vestigate this hypothesis. The dependent variable in the GLMM was
preference for savory foods or fruit tastiness ratings. A temperature
dummy variable (warm = 1, cool = 0), the perceived temperature of
savory foods, or the tastiness ratings of fruits, as well as their interac-
tion, were set as fixed factors. Participants were set as random factors.
When the interaction was significant, a post hoc analysis was con-
ducted, separating data into warm and cool conditions.
The results demonstrated that perceptions of the temperature of
savory foods, which were dependent on the ambient temperature, dif-
ferently influenced consumer preferences (B = -0.333, SE = 0.083,
p < 0.001; Fig. 4). When tastiness, healthfulness, and their interaction
with temperature (treated as a dummy variable) were additionally
controlled, the effect of perceptions of the temperature of savory foods
(B = −0.169, SE = 0.066, p = 0.010), but not of their perceived tas-
tiness (B = −0.089, SE = 0.068, p = 0.191) or healthfulness
(B = 0.027, SE = 0.068, p = 0.689), on food preferences differed ac-
cording to the ambient temperature. The effect of ambient temperature
on perceptions of the temperature of fruits was not significant
(B = −0.129, SE = 0.095, p = 0.177). The perceived temperature of
fruits influenced fruit tastiness ratings regardless of whether partici-
pants were in warm or cool ambient temperature.
Post hoc analysis revealed that the perceived food temperature of
savory foods decreased preferences for these foods in a warm ambient
temperature (B = −0.205, SE = 0.058, p < 0.001). After tastiness and
healthfulness were controlled, the effect of perceived food temperature
remained the same (B = −0.115, SE = 0.047, p = 0.016). In a cool
ambient temperature, the perceived food temperature increased pre-
ferences for these foods (B = 0.129, SE = 0.060, p = 0.033). However,
after tastiness and healthfulness were controlled, the effect of perceived
food temperature disappeared (B = 0.054, SE = 0.047, p = 0.254).
These results indicate that the reduced preference for savory foods
in warm ambient temperature derived from the perceived temperature
of these foods.
Similar results were obtained in all analyses when age and sex were
included as covariates.
K. Motoki et al.
3.6. Item analysis: does perceived food temperature contribute to different
preferences for savory foods under warm and cool ambient conditions?
The set of 10 savory foods (Appendix Fig. A1) included a variety of
images, which were perceived as relatively warm or cool (Appendix
Table A1). We investigated whether perceived temperature was related
to differences in the preferences for each food under warm and cool
ambient conditions. Spearman’s rank-order correlation test revealed
that perceived food temperature was negatively correlated with dif-
ferences in the preferences for savory foods under the warm and cool
ambient conditions (r = −0.770, p = 0.009; Fig. 5).
These results indicate that perceived food temperature contributed
to differences in savory food preferences in warm and cool ambient
temperatures.
4. Discussion
Grocery stores sell a variety of food and may use air-conditioning to
manage ambient temperature. Different theories have yielded different
hypotheses about the effect of ambient temperature on preferences for
different categories of food images. Embodiment theory proposes that
warm ambient temperature increases food preferences regardless of
category (H1). Thermoregulation theory proposes that, given the role of
food intake in regulating body temperature, warm ambient temperature
may decrease food preferences regardless of category (H2a). Combining
notions about perceived food temperature with thermoregulation
theory, we hypothesized that warm ambient temperature may influence
food preferences as a function of perceived food temperature, which, in
turn, depends on its category (H2b). Results of the study demonstrated
that the effect of warm ambient temperature on food preferences is
dependent on food category (partially supporting H2b). Participants in
warm ambient temperature displayed decreased preferences for savory
foods, and this effect was modulated by the perceived temperature of
savory foods. These findings are the first evidence of the effect of warm
ambient temperature on preferences for different categories of food and
provide theoretical and practical insights into how sensory input in-
fluences consumers’ food evaluations.
6
Food Quality and Preference 69 (2018) 1–9
Fig. 4. Differential influence of perceived food
warmth on preference for savory foods in warm and
cool temperatures. Differential influence of perceived
warmth on preferences for savory foods in warm
versus cool temperatures. In warm temperatures, the
perceived warmth of savory foods negatively influ-
enced preferences for savory foods. In cool tempera-
tures, it positively influenced preferences for savory
foods. Gray shadow indicates 95% confidence inter-
vals.
4.1. Theoretical contribution
The findings of this study contribute to the understanding of ther-
moregulatory effects on food preferences by considering perceived food
temperature. Classic thermoregulation theory focuses on the role of
food intake in regulating body temperature but does not explicitly ac-
count for the effect of food temperature (Johnson & Kark, 1947). Per-
ceived food temperature differs according to category of food. In gen-
eral, savory foods were perceived to be warmer than foods in other
categories (sweets, vegetables, and fruits). The intake of warm food
may be an inefficient way to reduce body temperature. Our research
design enabled us to determine that the preference for food decreases
only for savory (warm) foods in warm ambient temperature. Ad-
ditionally, the perceived temperature of savory food modulated the
effect of ambient temperature on consumer preferences, whereas other
factors that contribute to food preferences (tastiness and healthfulness)
did not. These findings increase our understanding of the relevance of
thermoregulation theory to food preference and indicate that only
savory foods are devalued in warm ambient temperature.
Contrary to our hypothesis, ambient warm temperature did not in-
crease preferences for fruits. According to thermoregulation theory,
food intake is used to increase body temperature. A representative
study demonstrated that humans consume more food during prolonged
cool days to increase their body temperature (Johnson & Kark, 1947).
Among food categories, savory foods can be used to increase body
temperature, whereas it appears that fruits do not serve this purpose.
Thus, food intake may be used only to increase body temperature.
Additionally, fruit is not cooler than iced or frozen foods, such as ice
cream. Although fruit was perceived as cooler than foods in other ca-
tegories, the images of fruit used in this study depicted them as raw and
not frozen. Raw fruit might not be the first choice for reducing body
temperature in daily life. These two non-exclusive explanations for why
warm ambient temperature did not increase fruits preferences require
further testing.
The results contribute to a greater understanding of the effect of
warmth on perception and preference of both food and non-food pro-
ducts. The effects of room temperature on consumer preferences are
K. Motoki et al.
small but significant. A previous study found that ambient warmth in-
creases product preferences by about 10% (Zwebner et al., 2013) while
in the present study, ambient warmth decreased preferences for savory
foods by about 8%. Although the size of the effect is not large, our
follow-up analyses confirmed that the effect was derived from the
perceived warmth of savory food. When savory food was perceived to
be warmer, participants experienced a more decreased preference for
savory food. This finding partially explains why the effect of ambient
warmth on preferences is generally not large, and also indicates when
this effect might be strengthened or weakened. In fact, the study re-
ferred to above found that ambient warmth increases product valuation
for some products possibly perceived as cool (e.g., bath gel, milk) but
not for others (e.g., batteries; Zwebner et al., 2013). In this context, the
findings of the present study indicate the existence of an effect of am-
bient warmth on preferences for savory food, and also clarify the un-
derlying mechanisms that determine when such effects are stronger or
weaker.
4.2. Practical contribution
Managers of grocery stores and restaurants can readily control
ambient temperature. Indeed, ambient temperature is a subtle en-
vironmental factor, and modifying it is less likely to lead to negative
consumer reactions. Marketers may be able to influence food pre-
ferences by varying the ambient temperature. The results of this study
demonstrate that ambient warmth does not influence all categories of
food; indeed, the effect was limited to savory foods. In addition, the
effect was amplified in warm savory foods. Based on these findings,
store managers should be primarily concerned about the effect of am-
bient temperature on sales of warm savory foods. Store managers can
utilize the relationship between ambient warmth and preferences for
warm savory foods by reducing ambient warmth in the warm savory
foods section.
4.3. Limitations and future research
Because this study evaluated preferences for food images, it is not
known how the results relate to actual foods. Indeed, differences be-
tween actual foods and images thereof render studies that use the latter
of limited generalizability to real-life situations that involve the former,
7
Food Quality and Preference 69 (2018) 1–9
Fig. 5. Item analysis: the perceived warmth of savory
foods was negatively correlated with preferences for
such foods. Food temperature contributes to different
levels of preference for savory foods under warm
versus cool ambient conditions. The perceived
warmth of savory foods was negatively correlated
with the warmth of the ambient temperature. Gray
shadow indicates 95% confidence intervals.
as the actual foods induce more intense sensory experiences (e.g., smell,
taste, touch) and allow for measurement of food intake.
Previous studies have shown that the aroma of savory foods (e.g.,
cheese soup, cheddar cheese) is more intense in warm environments
(Drake, Yates, & Gerard, 2005; Kähkönen, Tuorila, & Hyvönen, 1995).
Moreover, savory foods (e.g., chicken broth and miso soup) were rated
as less salty and less preferred in warm environments (Kim, Samant,
Seo, & Seo, 2015). Further research (e.g., a similar experiment in
winter) is needed to confirm whether ambient warm temperature in-
fluences actual taste experiences (e.g., aroma, taste), preferences, and
the intake of savory warm foods.
It is difficult to conclude whether the effects in this study were
driven by absolute or relative temperatures. There has been some re-
cent discussion of whether the effects of sensory information are ab-
solute or relative (Sunaga, 2018); a recent study showed that the effects
of auditory/visual senses on cognitive performance are driven by re-
lative rather than absolute stimuli (Brunetti, Indraccolo, Del Gatto,
Spence, & Santangelo, 2018). Further study is needed to clarify whether
the effect of warmth is driven by absolute or relative temperature. If the
effect is relative, is it derived from temperature changes between out-
door and indoor environments, or between indoor rooms with different
temperatures? These issues require further investigation.
Emotions associated with disgust resulting from the idea of food
poisoning may be another factor underlying the effect of warmth on the
decreased preference for savory foods. Warm temperatures generally
increase microbial populations (i.e., the source of poisoning) in savory
foods (Hobbs & Roberts, 1987). Incidental disgust reduces the pre-
ference for food (Motoki & Sugiura, 2018). Thus, consumers may per-
ceive a higher risk of food poisoning and/or feel disgust in relation to
savory foods in warm conditions. These factors may explain the effect of
warmth on savory food preferences and require further investigation.
Future research should also investigate how sensory interactions
influence eating behaviors. Although many studies on ambient tem-
perature have focused on a single sense, different senses interact, and
this interaction influences eating behaviors (Spence, Puccinelli, Grewal,
& Roggeveen, 2014; Velasco, Woods, Petit, Cheok, & Spence, 2016). It
has been shown that bright light enhances the perception of warmth
and increases preferences for spicy food as well as orange juice (Xu &
Labroo, 2014), which seems inconsistent with the present finding. The
interaction of warmth and visual (ambient brightness) or gustatory
K. Motoki et al. Food Quality and Preference 69 (2018) 1–9

(spicy, sweet) experiences may influence food preferences. Thus, future
research should address how multisensory experiences influence eating
behavior.
Although our use of a variety of food items may render our results
generalizable, our findings should be confirmed by additional research.
However, replication of this study with actual foods may be difficult
because we used a variety of food images (10 for each food category),
which may render measurements of actual consumption problematic.
Nonetheless, we could not include all the relevant food items in each
category and did not include beverages (e.g., coffee, tea). Given that
beverages are potentially warmer/cooler than foods, warm/cool drinks
might affect consumer preferences in cool/warm temperature, regard-
less of taste category.
These experiments were conducted during the early summer, which
may have affected the results. Estimates of food intake differ in summer
and winter (Capita & Alonso-Calleja, 2005; Joachim, 1997), and the
effects of ambient temperature on food preferences may be influenced
by season. For example, the reduced preference for savory warm foods
may increase in summer. Further research (e.g., a similar experiment in
winter) is needed to confirm whether these finding are generalizable.
5. Conclusion
This study investigated alternative hypotheses regarding the effects
of ambient temperature on preferences for different categories of food
images. The results partially support the efficient thermoregulation
hypothesis, demonstrating that ambient warm temperature decreases
preferences for savory foods. Indeed, perceived food temperature, but
not tastiness or healthfulness, modulated the decreased preferences for
savory foods. The efficient thermoregulation hypothesis was supported
only for warm savory foods and not for fruits. By incorporating the
perceived food temperature category into thermoregulation theory, this
study improved our understanding of how ambient temperature influ-
ences consumer food preferences.
Acknowledgments
The authors would like to thank the anonymous reviewers for their
helpful comments, which greatly improved the final version of the
paper. This study was supported by JSPS KAKENHI Grant Number
17J00389 (K.M.), KAKENHI Grant Number 16H01873 from MEXT
(M.S.), and a Grant-in-Aid for Scientific Research on Innovative Areas
(Research in a proposed research area: 17H06046) (R.N.).
The English in this document has been checked by at least two
professional editors, both native speakers of English. For a certificate,
please see: http://www.textcheck.com/certificate/vZfvAq.

Appendix

Fig. A1. List of savory foods.

Table A1
Comparison of savory food preferences under warm and cool ambient temperature conditions.
Savory food preference Food warmth (warm and cool Preference (warm room) Preference (cool room) Preferences (warm–cool P value (preference: warm–cool)
room) room)

Pizza a 6.06 4.85 5.36 −0.51 0.104

Pizza b 5.94 4.59 5.44 −0.85 0.017*
Pizza c 6.17 4.59 5.52 −0.93 0.010*
Potato snack 4.38 4.56 4.68 −0.12 0.760
Boiled shrimp 3.40 5.26 5.64 −0.38 0.303
Sushi a 3.29 6.19 6.20 −0.01 0.956
Harm and sausage 3.62 4.19 4.40 −0.21 0.594
Sushi b 3.42 4.52 4.40 0.12 0.760
Fried potatoes 5.85 4.56 4.88 −0.32 0.396
Smoked salmon 5.08 4.96 5.56 −0.60 0.053

*
p < 0.05.

8
K. Motoki et al.
References
Baron, R. A., & Bell, P. A. (1976). Aggression and heat: The influence of ambient tem-
perature, negative affect, and a cooling drink on physical aggression. Journal of
Personality and Social Psychology, 33(3), 245.
Barsalou, L. W. (2010). Grounded cognition: Past, present, and future. Topics in Cognitive
Science, 2(4), 716–724.
Bates, D., Mächler, M., Bolker, B., Walker, S. (2014). Fitting linear mixed-effects models
using lme4. arXiv preprint arXiv:1406.5823.
Bell, P. A., & Baron, R. A. (1977). Aggression and ambient temperature: The facilitating
and inhibiting effects of hot and cold environments. Bulletin of the Psychonomic
Society, 9(6), 443–445.
Blechert, J., Meule, A., Busch, N. A., & Ohla, K. (2014). Food-pics: An image database for
experimental research on eating and appetite. Frontiers in Psychology, 5, 617.
Brobeck, J. R. (1948). Food intake as a mechanism of temperature regulation. The Yale
Journal of Biology and Medicine, 20(6), 545.
Brunetti, R., Indraccolo, A., Del Gatto, C., Spence, C., & Santangelo, V. (2018). Are
crossmodal correspondences relative or absolute? Sequential effects on speeded
classification. Attention, Perception, & Psychophysics, 80(2), 527–534.
Capita, R., & Alonso-Calleja, C. (2005). Differences in reported winter and summer
dietary intakes in young adults in Spain. International Journal of Food Sciences and
Nutrition, 56(6), 431–443.
Drake, M., Yates, M., & Gerard, P. (2005). Impact of serving temperature on trained panel
perception of Cheddar cheese flavor attributes. Journal of Sensory Studies, 20(2),
147–155.
Epstein, L. H., Carr, K. A., Scheid, J. L., Gebre, E., O'Brien, A., Paluch, R. A., & Temple, J.
L. (2015). Taste and food reinforcement in non-overweight youth. Appetite, 91,
226–232.
Hare, T. A., Camerer, C. F., & Rangel, A. (2009). Self-control in decision-making involves
modulation of the vmPFC valuation system. Science, 324(5927), 646–648.
Hobbs, B. C., & Roberts, D. (1987). Food poisoning and food hygiene (5th ed.). London:
Edward Arnold.
Howard, J. D., Gottfried, J. A., Tobler, P. N., & Kahnt, T. (2015). Identity-specific coding
of future rewards in the human orbitofrontal cortex. Proceedings of the National
Academy of Sciences, 112(16), 5195–5200.
Huang, X. I., Zhang, M., Hui, M. K., & Wyer, R. S. (2014). Warmth and conformity: The
effects of ambient temperature on product preferences and financial decisions.
Journal of Consumer Psychology, 24(2), 241–250.
IJzerman, H., Coan, J. A., Wagemans, F., Missler, M. A., Beest, I. V., Lindenberg, S., &
Tops, M. (2015). A theory of social thermoregulation in human primates. Frontiers in
Psychology, 6, 464.
IJzerman, H., & Semin, G. R. (2009). The thermometer of social relations: Mapping social
proximity on temperature. Psychological Science, 20(10), 1214–1220.
Joachim, G. (1997). The influence of time on dietary data: Differences in reported
summer and winter food consumption. Nutrition and Health, 12(1), 33–43.
Johnson, R. E., & Kark, R. M. (1947). Environment and food intake in man. Science
(Washington), 378–379.
Kähkönen, P., Tuorila, H., & Hyvönen, L. (1995). Dairy fat content and serving tem-
perature as determinants of sensory and hedonic characteristics in cheese soup. Food
Quality and Preference, 6(2), 127–133.
Keast, R. S., & Breslin, P. A. (2003). An overview of binary taste–taste interactions. Food
Food Quality and Preference 69 (2018) 1–9
Quality and Preference, 14(2), 111–124.
Kim, J.-W., Samant, S. S., Seo, Y., & Seo, H.-S. (2015). Variation in saltiness perception of
soup with respect to soup serving temperature and consumer dietary habits. Appetite,
84, 73–78.
Krishna, A. (2012). An integrative review of sensory marketing: Engaging the senses to
affect perception, judgment and behavior. Journal of Consumer Psychology, 22(3),
332–351.
Krishna, A., Morrin, M., & Sayin, E. (2013). Smellizing cookies and salivating: A focus on
olfactory imagery. Journal of Consumer Research, 41(1), 18–34.
Krishna, A., & Schwarz, N. (2014). Sensory marketing, embodiment, and grounded cog-
nition: A review and introduction. Journal of Consumer Psychology, 24(2), 159–168.
Motoki, K., Saito, T., Nouchi, R., Kawashima, R., & Sugiura, M. (2018). Tastiness but not
healthfulness captures automatic visual attention: Preliminary evidence from an eye-
tracking study. Food Quality and Preference, 64, 148–153.
Motoki, K., & Sugiura, M. (2018). Disgust, sadness, and appraisal: Disgusted consumers
dislike food more than sad ones. Frontiers in Psychology, 9, 76.
Nicol, A., & Young, B. (1989). Effects of the ingestion of warm, cold and frozen food on
thermal balance in cattle. Livestock Production Science, 23(1–2), 107–115.
North, A. C., Hargreaves, D. J., & McKendrick, J. (1999). The influence of in-store music
on wine selections. Journal of Applied Psychology, 84(2), 271.
Olsen, A., Ritz, C., Hartvig, D. L., & Møller, P. (2011). Comparison of sensory specific
satiety and sensory specific desires to eat in children and adults. Appetite, 57(1), 6–13.
Padulo, C., Carlucci, L., Manippa, V., Marzoli, D., Saggino, A., Tommasi, L., ... Brancucci,
A. (2017). Valence, familiarity and arousal of different foods in relation to age, sex
and weight. Food Quality and Preference, 57, 104–113.
Peirce, J. W. (2007). PsychoPy—Psychophysics software in Python. Journal of
Neuroscience Methods, 162(1), 8–13.
Sørensen, L. B., Møller, P., Flint, A., Martens, M., & Raben, A. (2003). Effect of sensory
perception of foods on appetite and food intake: A review of studies on humans.
International Journal of Obesity, 27(10), 1152.
Spence, C. (2012). Managing sensory expectations concerning products and brands:
Capitalizing on the potential of sound and shape symbolism. Journal of Consumer
Psychology, 22(1), 37–54.
Spence, C., Puccinelli, N. M., Grewal, D., & Roggeveen, A. L. (2014). Store atmospherics:
A multisensory perspective. Psychology & Marketing, 31(7), 472–488.
Sunaga, T. (2018). How the sound frequency of background music influences consumers’
perceptions and decision making. Psychology & Marketing, 35(4), 253–267.
Terrien, J., Perret, M., & Aujard, F. (2011). Behavioral thermoregulation in mammals: A
review. Frontiers in Bioscience, 16, 1428–1444.
Velasco, C., Woods, A. T., Petit, O., Cheok, A. D., & Spence, C. (2016). Crossmodal cor-
respondences between taste and shape, and their implications for product packaging:
A review. Food Quality and Preference, 52, 17–26.
Williams, L. E., & Bargh, J. A. (2008). Experiencing physical warmth promotes inter-
personal warmth. Science, 322(5901), 606–607.
Xu, A. J., & Labroo, A. A. (2014). Incandescent affect: Turning on the hot emotional
system with bright light. Journal of Consumer Psychology, 24(2), 207–216.
Zhong, C.-B., & Leonardelli, G. J. (2008). Cold and lonely: Does social exclusion literally
feel cold? Psychological Science, 19(9), 838–842.
Zwebner, Y., Lee, L., & Goldenberg, J. (2013). The temperature premium: Warm tem-
peratures increase product valuation. Journal of Consumer Psychology, 24(2),
251–259.