Content uploaded by Syllas Borburema Silva Oliveira
Author content
All content in this area was uploaded by Syllas Borburema Silva Oliveira on Feb 13, 2018
Content may be subject to copyright.
Original article
The influence of beverages on residual spiciness elicited by eating
spicy chicken meat: time-intensity analysis
Shilpa S. Samant, Sungeun Cho, Andrew D. Whitmore, Syllas B. S. Oliveira, Thais B. Mariz & Han-Seok Seo*
Department of Food Science, University of Arkansas, 2650 North Young Avenue, Fayetteville, AR 72704, USA
(Received 22 May 2016; Accepted in revised form 5 July 2016)
Summary This study aimed to determine whether five types of beverage, milk (3.3% fat), oolong tea, tomato juice,
sparkling water and spring water, reduce the residual spiciness elicited by eating spicy chicken over time.
After tasting a piece of spicy popcorn chicken, participants were asked to drink one of five beverages one
after another and rate the spiciness intensity using the time-intensity (TI) analysis. Based on the TI
parameters, milk was found to be the best beverage to reduce residual spiciness of spicy chicken over
time. Participants rated milk the most effective in decreasing residual spiciness of spicy chicken. Partial
least squares regression revealed that the three TI parameters, area under the curve, decreasing area and
increasing area, are the best predictors of the self-reported ratings for spiciness-reduction effect of individ-
ual beverages. In conclusion, this study shows that milk (3.3% fat) can be used as a palate cleanser for
spicy chicken.
Keywords Beverage, milk, palate cleanser, spiciness, spicy food, time-intensity analysis.
Introduction
Spicy food has been a dynamic area of research for a
long time owing to its popularity across different cul-
tures. In general, capsaicin (from red peppers) and
piperine (from black peppers) are the major chemical
constituents responsible for the ‘burning’ sensation
experienced in the mouth due to most spicy foods. The
burning sensation elicited by spicy food consumption
is referred with many terms including ‘burn’, ‘oral
pain’ and ‘irritation’. Sensorial liking and acceptance
of spicy food have been found to vary among individ-
uals. Some consumers have a high desirability for
spicy food (‘likers’), whereas other consumers almost
averse it (‘nonlikers’) (Tepper et al., 2004). A major
reason for this division between ‘likers’ and ‘nonlikers’
might be tolerance and liking of the spiciness in the
mouth, which in turn depends upon factors such as
gender, culture, personality and prior exposure (Ludy
& Mattes, 2012; Byrnes & Hayes, 2015).
Spiciness, in addition to sensations such as burn and
irritation as mentioned above, can cause deterioration
of other sensory attribute perceptions (Lawless et al.,
1985; Simmons et al., 2002). Specifically, Simmons
et al. (2002) found that sweet taste intensity in humans
decreased if the tongue was pretreated with capsaicin
solution. Moreover, it has been found that spiciness
perception is persistent over a period of time even
after the spicy food has been swallowed or expecto-
rated (Hutchinson et al., 1990). As a result, the resid-
ual spiciness in the mouth can modulate sensory
perception and liking of foods that are subsequently
consumed.
From a sensory analysis standpoint, residual spici-
ness complicates sensory perception of spicy food due
to its strong carry-over effect leading to early fatigue
(Cartwright et al., 1952). Researchers generally use the
terms, sensitisation and desensitisation, for this pur-
pose. Sensitisation is referred to as the phenomenon of
increasing oral irritation with successive spicy food
samples. However, if a gap of few minutes is allowed
between the spicy food samples, this irritation over
successive samples was found to be lower; such a phe-
nomenon is referred to as desensitisation (Prescott,
1999). Nature that sensitisation and desensitisation of
the tongue differs among individuals may affect sen-
sory analysis of multiple spicy foods. It has been
advised to conduct sensory analysis of spicy food with
only limited number of samples (e.g., 3–4) to avoid
influences of strong residual spiciness and sensory fati-
gue (Cartwright et al., 1952). Generally, palate cleans-
ers are used to avoid carry-over effect between
consecutive samples. However, it has been a challenge
for researchers to develop efficient palate cleansers for
*Correspondent: Fax: +1 479 575-6936; E-mail:hanseok@uark.edu
International Journal of Food Science and Technology 2016, 51, 2406–2415
doi:10.1111/ijfs.13221
©2016 Institute of Food Science and Technology
2406
spicy foods due to the strength of carry-over effect. To
overcome these problems from a consumer as well as
research standpoint, it is important to minimise the
carry-over effect or residual spiciness in the mouth
experienced during spicy food consumption.
In 1986, Stevens and Lawless studied the impact of
four basic taste solutions on irritation of capsaicin and
piperine solutions on the tongue using magnitude esti-
mation method. Sucrose and citric acid solutions were
found to mitigate the irritation faster than other taste
solutions. In addition, citric acid was found to show
stronger reduction in capsaicin irritation compared to
piperine. Apart from basic taste solution, certain bev-
erages, such as milk, have also been reported to reduce
residual spiciness perception in the mouth. More
specifically, Kitsawad & Nguyen (2014) investigated
water, bread, unsweetened milk and sucrose solution
as potential palate cleansers for spicy tom yum soup
based on spiciness ratings on a 10-point intensity scale.
They found that milk was the best choice to reduce
spiciness of the soup. However, this notion is still
under debate and needs further scientific evidence to
support it (Nasrawi & Pangborn, 1990; Lucak & Del-
wiche, 2009). For instance, Lucak & Delwiche (2009)
investigated effect of milk as a potential palate cleanser
for spicy tortilla chips along with table water and
crackers using a 9-point rating scale. The results
showed that even though whole milk did work well to
reduce burn/heat intensity it did not provide any
advantage over the traditional table water and crack-
ers. Moreover, table water and crackers performed
consistently over replicates, unlike milk that showed
inconsistency in reducing spiciness perception. There-
fore, it cannot be said with affirmation whether milk is
a good beverage option to reduce residual spiciness in
the mouth and whether it is the fat in milk that is
responsible for this reduction.
Other beverage characteristics such as temperature
and carbonation can also influence spiciness percep-
tion. For example, spiciness has a property of being
intensified at warmer temperatures and inhibited at
cooler ones (Green, 1986; Reinbach et al., 2009).
Carbonation has not been explored much with
respect to its impact on spiciness perception. Beer is
suggested as a complimentary beverage with spicy
foods by food experts (Dornenburg & Page, 2009).
However, Cartsens et al. (2002) reported that spicy
food and carbonation are both considered as ‘irri-
tants’ that can cause burning or stinging sensation
on the tongue. This means that their combination
might actually worsen oral pungency. Because exact
mechanism of how either temperature or carbonation
affects spiciness perception is still unknown, further
research in this regard is essential to provide conclu-
sive evidence.
One of the limitations of research done so far to
evaluate the effect of different beverages on spiciness
perception is that many studies did not take the ‘time’
factor into account. However, the time factor should
not be ignored because spiciness perception is not a
momentary sensation and generally lingers in the
mouth for a period of time, even after swallowing or
expectorating (Hutchinson et al., 1990; Cliff & Hey-
mann, 1993). It seems that the time-intensity (TI) anal-
ysis is an efficient methodology for evaluating changes
in spiciness intensity over time. The TI analysis is a
descriptive analysis method in which participants with
adequate training rate their perceived sensations of a
specific sensory attribute over a period of time. The
parameters drawn from the TI curve give a reliable
estimate of intensity variations with time (Duizer
et al., 1995; Meilgaard et al., 2006). The TI analysis
has been used to evaluate changes in sensory attribute
intensity over time in a variety of food and beverage
products including bread (De Morais et al., 2013), ice
creams (Cadena & Bolini, 2011), chocolate milk for-
mulation added with chia oil (Rodrigues et al., 2015),
raspberry-flavoured gelatin (Palazzo & Bolini, 2009)
and potato chips (Luckett et al., 2016).
Based on the above-mentioned information, a pri-
mary objective of this study was to determine the
effect of different beverages on residual spiciness eli-
cited by eating spicy food. The beverages chosen for
the study were milk, oolong tea, tomato juice, spark-
ling water and spring water based on preliminary tests
with common beverages. Milk was chosen as a bever-
age sample because it has shown potential to be effec-
tive in mitigating spiciness (Nasrawi & Pangborn,
1990; Lee & Kim, 2013; Lucas & Delwiche, 2009).
Because sucrose and citric acid solutions have been
found to reduce capsaicin burn (Stevens & Lawless,
1986), tomato juice (which contains both compounds)
was chosen. Because oolong tea has been considered
as a good beverage complement to spicy food by some
food experts (Dornenburg & Page, 2009), it was
included as a test sample. Sparkling water was chosen
as a representative carbonated beverage to examine
the effect of carbonation on spiciness reduction.
Finally, plain water was included as the control sam-
ple. Considering the time function of spiciness percep-
tion, the reduction effect of these beverages on residual
spiciness elicited by spicy chicken consumption was
evaluated over time. Among spicy foods, spicy chicken
meat was chosen due to its popularity in the U.S.A.
Materials and methods
This study was conducted in accordance with the Dec-
laration of Helsinki for studies on human participants.
The protocol used in this study was approved by the
Institutional Review Board of the University of
©2016 Institute of Food Science and Technology International Journal of Food Science and Technology 2016
Palate cleanser for spicy chicken meat S. S. Samant et al. 2407
Arkansas (Fayetteville, AR, USA). After explaining
the experimental procedure, a written informed con-
sent was obtained from each participant prior to the
study.
Participants
Thirty-six healthy adults (16 men and 20 women;
mean age standard deviation =36 8 years) were
recruited through a consumer profile database from
the University of Arkansas Sensory Service Center
(Fayetteville, AR, USA) that contains information
about more than 6200 Northwest Arkansas residents.
Participants were screened prior to selection based on
an online screener that asked questions about their
spicy food consumption frequency and liking. A mini-
mum frequency of 1–2 times per week and a minimum
liking of 7 on a 9-point hedonic scale were considered
as the criteria for selection. Such criteria were to avoid
bias due to nonliking or nonconsumption of spicy
food (Ludy & Mattes, 2012). Additionally, to avoid
cultural bias, only Caucasian participants were chosen.
Samples
A popular spicy food sample, popcorn chicken (Great
Value
TM
, Wal-mart Stores, Inc., Bentonville, AR,
USA) with hot sauce (Buffalo Wild Wings, Inc.,
Columbus, OH, USA) was chosen for this study
(henceforth to be referred as spicy chicken). Uniform-
sized frozen popcorn chicken pieces were baked in a
preheated oven at 190.6 °C for 11 min and allowed to
further sit in the oven for additional 2 min (in accor-
dance with package instructions). A preweighed
amount of hot sauce (45 g/30 pieces) was mixed with
the baked popcorn chicken. This amount of hot sauce
was chosen based on preliminary studies performed by
the authors. This level of spiciness was found to be
just enough to cause the desired level of spicy burn in
the mouth.
Sample preparation
The beverage samples chosen were Vitamin D Milk
(3.3% fat) (Great Value
TM
, Wal-mart Stores, Inc.),
Pouchong Oolong Tea (DavidsTea, Montr
eal, Qu
ebec,
Canada), Tomato Juice (Great Value
TM
, Wal-mart
Stores, Inc.), Sparkling Water (Ozarka
Ò
, Nestl
e
Waters North America Inc., Stanford, CT, USA) and
Spring Water (Mountain Valley Springs Co., LLC Hot
Springs, AR, USA) as a control. No beverage, except
for oolong tea, required any preparation before serv-
ing. Oolong tea was prepared in water at 76.7 °C
(10 g tea/475-mL water) for 5 min in accordance with
package instructions. It was allowed to cool to 23 °C
prior to serving. To minimise any potential effect of
serving temperature on spiciness perception, all bever-
age samples were served at room temperature (approx-
imately, 23 °C) in white souffl
es cups (118-mL).
Procedure
Overview
This was a 3-day study including one training session
and two test sessions. Each session was conducted on
a separate day and each day was a week apart. The
detailed methodology of TI analysis as well as implica-
tions of each session is given below.
Training session
On Day-1, each participant was asked to participate in
a training session. As mentioned earlier, the TI analy-
sis is a descriptive analysis conducted with trained
panellists. However, it is slightly different from con-
ventional descriptive analysis techniques with respect
to panel recruitment and training. In many studies
using the TI analysis, na€
ıve consumers have been
recruited and trained prior to the actual experiment
(Reinbach et al., 2009; Nasrawi & Pangborn, 1990).
To obtain reliable results, participants were trained
according to the method inspired from Peyvieux &
Dijksterhius (2001). First, participants were given a
brief overview of the study. Paper version of TI analy-
sis was provided to the participants to explain the
methodology. The term ‘spiciness’ was made clear to
the participants to avoid bias due to other sensory
attributes. Once they were familiarised with the proce-
dure (detailed explanation in section 2.4.4.), individual
in-booth training with ‘dummy’ beverage samples was
completed. A carbonated beverage (Coca Cola
Ò
, The
Coca Cola Company, Atlanta, GA, USA) and apple
juice (Great Value
TM
, Wal-mart Stores, Inc.) were
selected as ‘dummy’ beverages to avoid any expecta-
tion error from the participants during actual test ses-
sions.
Test sessions
Day-2 and Day-3 were test sessions. During each test
session, spring water was provided as a control sam-
ple, along with two of the remaining four beverage
samples (milk, tomato juice, oolong tea and sparkling
water). The order of the beverages was randomised
across both days ensuring spring water was presented
on both days. The purpose of dividing samples across
two session days was to limit the number of samples
per session. According to previous research, there is a
high carry-over effect of spiciness if more than 3–4
samples are used at one time (Cartwright et al., 1952).
Time-intensity analysis
To measure temporal dynamics of spiciness perception,
the time-intensity analysis was used using the sensory
©2016 Institute of Food Science and TechnologyInternational Journal of Food Science and Technology 2016
Palate cleanser for spicy chicken meat S. S. Samant et al.2408
analysis software, Compusense
Ò
five (Release 5.6;
Compusense Inc., Guelph, ON, Canada). The scale
provided to the participants was a vertical visual ana-
logue scale. Participants were informed that the bot-
tom of the scale indicated lowest intensity (‘0’) and the
top of the scale indicated maximum intensity (‘100’).
The scale had a ‘start’ button at the bottom. Partici-
pants rinsed their mouth twice with spring water
(Mountain Valley Springs Co.) to cleanse their palates
before beginning; the water was expectorated after
rinsing and not swallowed. Each participant received
one spicy chicken per beverage sample. Participants
were asked to take the whole chicken sample in their
mouth and right-click (using the mouse) on the ‘start’
button. They could move the cursor (using the mouse)
up (indicating stronger spiciness) or down (indicating
lower spiciness). Participants were asked to chew the
food sample for around 30 s and then swallow. The
time of swallowing for each participant was kept
approximately the same because earlier research
showed differences in spicy perception before and after
swallowing (Hutchinson et al., 1990; Cliff & Heymann,
1993). Participants continued rating their perceived
spiciness for a total of 45 s. This was the considered
‘initial spiciness’ of the food sample.
Next, participants were asked to drink the entire
beverage sample provided to them and start rating the
perceived spiciness in their mouth immediately (as
described above). This was the ‘residual spiciness’
which was measured for 120 s. However, participants
were asked to move the cursor back to the bottom of
the scale when they no longer perceived any spiciness
(in case it was before the end of 120 s). The test proce-
dure was repeated for the next beverage sample. A
break of 120 s was given between the two consecutive
samples during which participants cleansed their
palates by rinsing their mouths twice with water. The
duration of break was set up based on a preliminary
study as well as previous research (Lucak & Delwiche,
2009).
A total of 10 TI parameters were used: (i) time at
the maximum intensity (T
max
; time to reach peak
intensity), (ii) maximum intensity (I
max
; the highest
intensity on TI record), (iii) Area under the curve
(AUC; total area under the time-intensity curve), (iv)
duration (Dur; time from onset to return to baseline),
(v) increasing angle (Inc angle; rate of increase (linear
fit) from the onset to peak intensity), (vi) Increasing
area (Inc area; area under the curve from onset to
peak intensity), (vii) Decreasing angle (Dec angle; rate
of decrease (linear fit) from initial declining point to
baseline), (viii) Decreasing area (Dec area; area under
the curve from declining point to baseline), (ix) Initial
delay (Int delay; time from ingestion until onset of
sensation) and (x) Initial intensity (Int inten; intensity
at the onset of sensation); the definitions of the TI
parameters were determined based on previous studies
(Lawless & Heymann, 2010; Luckett et al., 2016).
Subjective ratings: spiciness-reduction effect and hedonic
impression of beverage sample
Participants were asked to rate spiciness-reduction
effect and hedonic impression immediately after TI rat-
ing of each beverage sample. First, participants were
asked to answer the question, ‘How effective/ineffective
do you think is this beverage sample to reduce the resid-
ual spiciness in your mouth?’ on a 9-point scale ranging
1 (extremely ineffective) to 9 (extremely effective). Sec-
ond, participants were asked to answer the question,
‘How much do you like/dislike this beverage sample with
the spicy food?’, on a 9-point hedonic scale ranging 1
(dislike extremely) to 9 (like extremely).
Data analysis
Data were analysed using JMP
Ò
Pro (version 12.0,
SAS Institute Inc., Cary, NC, USA). Ten TI parame-
ters and two subjective ratings (spiciness-reduction
effect and hedonic impression of beverage samples)
were collected using Compusense software. Results for
the spring water sample, in terms of TI parameters
and subjective ratings, on both days did not signifi-
cantly differ (P>0.05, for all). Thus, TI parameters
and subjective ratings for both spring water samples
were averaged for further analysis. A two-way analysis
of variance (ANOVA) was performed treating ‘bever-
age sample’ as a fixed effect and ‘panel’ as a random
effect. If a significant difference in means was indicated
by the ANOVA, post-hoc comparisons between inde-
pendent variables were performed using Tukey’s Hon-
est Significant Difference (HSD) method. In addition,
Partial Least Squares Regression (PLSR) was per-
formed to model self-reported spiciness-reduction
effect based on the ten TI parameters. PLSR is a
widely popular statistical approach used in sensory
and consumer studies to predict self-reported measures
such as overall impression based on descriptive sensory
attributes and/or physicochemical parameters (Cadena
et al., 2012, 2013). In this study, self-reported spici-
ness-reduction effect was treated as the dependent vari-
able (Y-matrix) and ten TI parameters were the
independent variables (X-matrix). Centred and scaled
values of the dependent variable (self-reported spici-
ness-reduction effect) and independent variables (10 TI
parameters) were used with Leave-One-Out cross-vali-
dation method to determine the lowest number of fac-
tors required to minimise Root Mean PRESS statistic
(RM-PRESS) and maximise percentage variation
explained for Y (Cox & Gaudard, 2013). The variable
influence on projection (VIP) values and model coeffi-
cients were used to determine which TI parameters (X)
were important in modelling the self-reported
©2016 Institute of Food Science and Technology International Journal of Food Science and Technology 2016
Palate cleanser for spicy chicken meat S. S. Samant et al. 2409
spiciness-reduction effect (Y). VIP values have been of
paramount interest in sensory and consumer studies as
they help in reducing a large set of independent vari-
ables to a smaller subset based on which variables are
important in predicting the dependent variable. Gener-
ally, independent variables with VIP values above 0.8
are considered to be important (Rossini et al., 2012).
In this study, weighted sums of squares of the PLS
weights were calculated from the Y-variance of each
PLS component to yield VIP values (Wold et al.,
2001). In addition, Pearson’s correlation analysis was
performed to determine the relationship between TI
parameters and self-reported spiciness-reduction effect.
A statistically significant difference was defined as
P<0.05.
Results
Four TI parameters (time at the maximum intensity,
maximum intensity, area under the curve and decreas-
ing area) are reported in the results as other parame-
ters (duration, increasing angle, increasing area,
decreasing angle, initial delay and initial intensity) did
not significantly differ among the five beverage condi-
tions (P>0.05, for all).
Initial spiciness
Prior to drinking beverage sample, ‘initial spiciness’
of spicy popcorn chicken was rated for 45 s. As
shown in Figs 1–4, the TI parameters for ‘initial spici-
ness’ were not significantly different among the five
beverage conditions (milk, oolong tea, sparkling
water, tomato juice and spring water) (P>0.05, for
all), indicating no significant difference in the initial
spiciness among the five beverage conditions. In
addition, no difference in the initial spiciness indicates
that there might be no significant difference among
the five beverage conditions with respect to the carry-
over effect of spiciness between consecutive beverage
conditions.
Residual spiciness
Following the ‘initial spiciness’ rating of spicy popcorn
chicken for 45 s, the ‘residual spiciness’ was assessed
immediately after drinking one of the five beverage
samples. Figure 5 graphically represents the mean
intensity rating of residual spiciness for each beverage
sample over a period of 120 s, with readings taken
every 0.5 s.
Figure 1 Comparison of the five beverages with respect to the time
at the maximum intensity before (initial spiciness) and after (residual
spiciness) drinking the beverages. N.S. represents no significant dif-
ference between beverages (P>0.05). Error bars represent standard
error of the means
Figure 2 Comparison of the five beverages with respect to the max-
imum intensity before (initial spiciness) and after (residual spiciness)
drinking the beverages. N.S. represents no significant difference
between beverages (P>0.05). *represents a significant difference
between beverages (P<0.05). Mean ratings with different small let-
ters indicate a significant difference between the beverages at
P<0.05. Error bars represent standard error of the means.
Figure 3 Comparison of the five beverages with respect to the area
under the curve before (initial spiciness) and after (residual spiciness)
drinking the beverages. N.S. represents no significant difference
between beverages (P>0.05). **represents a significant difference
between beverages (P<0.01). Mean ratings with different small let-
ters indicate a significant difference between the beverages at
P<0.05. Error bars represent standard error of the means.
©2016 Institute of Food Science and TechnologyInternational Journal of Food Science and Technology 2016
Palate cleanser for spicy chicken meat S. S. Samant et al.2410
Time at the maximum intensity of residual spiciness (T
max_
Res
)
T
max_Res
did not significantly differ among the five
beverage conditions [F(4, 140) =1.11, P=0.35]
(Fig. 1).
Maximum intensity of residual spiciness (I
max_Res
)
Beverage samples significantly influenced the maximum
intensity of residual spiciness [F(4, 140) =3.40,
P=0.01]. As shown in Fig. 2, maximum intensity of
residual spiciness was significantly lower when drank
milk compared to when drank oolong tea (P=0.03)
or sparkling water (P=0.02). Additionally, the
I
max_Res
for milk condition was marginally lower than
the I
max_Res
for spring water condition (P=0.05).
However, the I
max_Res
for milk condition was not
significantly different from the I
max_Res
for tomato
juice condition (P=0.07). Additionally, the I
max_Res
did not significantly differ among the four beverage
conditions: oolong tea, sparkling water, tomato juice
and spring water conditions (P>0.05 for all binary
combinations).
Area under the curve of the residual spiciness (AUC_
Res
)
As shown in Fig. 3, AUC_
Res
significantly differed
among beverage conditions [F(4, 140) =4.00,
P=0.004]. The AUC_
Res
for milk condition was sig-
nificantly lower than the AUC_
Res
for oolong tea
(P=0.01), sparkling water (P=0.01) and spring
water (P=0.02) conditions. However, the AUC_
Res
for tomato juice condition did not differ from the
AUC_
Res
for milk condition (P=0.12). In addition,
no significant differences were found among oolong
tea, sparkling water, tomato juice and spring water
conditions (P>0.05 for all binary combinations).
Decreasing area of residual spiciness (Dec area_
Res
)
As shown in Fig. 4, the decreasing area parameter of
the residual spiciness curve significantly differed among
the five beverage conditions [F(4, 140) =4.73,
P=0.001]. The Dec area_
Res
for milk condition was
significantly lower than the Dec area_
Res
for oolong
tea (P=0.002), sparkling water (P=0.03) and spring
water (P=0.006) conditions. However, no difference
was found between milk and tomato juice conditions
(P=0.30). In addition, no significant differences with
respect to the Dec area_
Res
were found among oolong
tea, sparkling water, tomato juice and spring water
conditions (P>0.05 for all binary combinations).
Spiciness-reduction effect of beverage samples
Five beverage samples’ spiciness-reduction effect was
found to be significantly different [F(4, 140) =17.58,
P<0.001]. Participants rated that milk was the most
effective in reducing residual spiciness compared to
oolong tea (P<0.001), sparkling water (P<0.001),
tomato juice (P<0.001) and spring water (P<0.001)
as shown in Fig. 6. Tomato juice showed a marginally
higher effect in mitigating residual spiciness than did
sparkling water (P=0.05). There were no significant
differences with respect to spiciness-reduction effect in
other binary comparisons (P>0.05).
Hedonic impression of beverage samples when consumed
with spicy chicken
As shown in Fig. 6, when consumed with spicy
chicken participants’ hedonic impression for five bever-
age samples was significantly different [F(4, 140)
=10.35, P<0.001]. Surprisingly, participants liked
spring water significantly more than oolong tea
Figure 5 Time-intensity curves of residual spiciness for the five bev-
erages: milk (3.3% fat), oolong tea, sparkling water, tomato juice
and spring water.
Figure 4 Comparison of the five beverages with respect to the
decreasing area before (initial spiciness) and after (residual spiciness)
drinking the beverages. N.S. represents no significant difference
between beverages (P>0.05). ** represents a significant difference
between beverages (P<0.01). Mean ratings with different small let-
ters indicate a significant difference between the beverages at
P<0.05. Error bars represent standard error of the means.
©2016 Institute of Food Science and Technology International Journal of Food Science and Technology 2016
Palate cleanser for spicy chicken meat S. S. Samant et al. 2411
(P<0.001), sparkling water (P<0.001) and tomato
juice (P=0.02); no significant difference was observed
between spring water and milk (P=0.72). Even milk
and tomato juice did not differ with respect to hedonic
impression when they were consumed with spicy
chicken (P=0.35). In addition, there were no signifi-
cant differences among tomato juice, oolong tea and
sparkling water in terms of hedonic impression when
consumed with spicy chicken (P>0.05 for all binary
combinations).
Relationship of the self-reported spiciness-reduction effect
with time-intensity parameters
Using Pearson’s correlation analysis and Partial Least
Squares regression (PLSR), the relationship of self-
reported spiciness-reduction effect with the ten TI
parameters was assessed. Table 1 shows the correlation
between the self-rated spiciness-reduction effect and
the TI parameters. Specifically, the self-rated spiciness-
reduction effect was significantly correlated with
T
max_Res
,(r=0.27, P<0.001), I
max_Res
(r=0.34,
P<0.001), Dur_
Res
,(r=0.31, P<0.001), AUC_
Res
(r=0.56, P<0.001), Inc area_
Res
(r=0.42,
P<0.001), Inc angle_
Res
(r=0.20, P=0.009) and
Dec area_
Res
(r=0.45, P<0.001). However, other
TI parameters such as Dec angle_
Res
(r=0.03,
P=0.73), Int delay_
Res
(r=0.04, P=0.62) and Int
inten_
Res
(r=0.09, P=0.25) did not show significant
correlations with subjective responses.
These results of correlation analysis were supported
by the PLSR model. PLSR analysis with Leave-one-
out cross-validation method determined that the PLSR
model with lowest RM-PRESS value (0.86) explained
31.72% variation in Y (self-rated spiciness-reduction
effect) using one latent vector. Six of ten TI param-
eters, namely T
max_Res
(b=0.08), I
max_Res
(b=
0.11), Dur_
Res
(b=0.10), AUC_
Res
(b=0.18),
Inc area_
Res
(b=0.13) and Dec area_
Res
(b=
0.14), were considered important variables in defin-
ing the final PLS model with each having variable
influence on projection (VIP) value above cut-off value
of 0.8 (Cox & Gaudard, 2013). Even though PLSR
model did not account for a great amount of variation
in the self-rated spiciness reduction, it shows a relative
impact of the TI parameters.
Discussion
The influence of beverages such as milk, oolong tea,
sparkling water, tomato juice and spring water on
residual spiciness of spicy popcorn chicken was evalu-
ated using the time-intensity (TI) methodology. Partic-
ipants were trained using the TI method and
instructed to rate their spiciness perception before
(‘initial spiciness’) and after (‘residual spiciness’) they
drank the beverage sample. TI curve parameters were
measured along with participants’ self-reported rating
of how effective they thought it was in reducing spici-
ness in their mouth and how much they liked the bev-
erage when consumed with the spicy chicken.
Overall, TI methodology seems to be an efficient
way to compare beverages in terms of their effect on
residual spiciness over time. Focusing on the TI curve
parameters, higher values of the time at the maximum
intensity (T
max
), the maximum intensity (I
max
), the
area under the curve (AUC), and the decreasing area
(Dec area) indicate more intense and persistent
Figure 6 Comparison of the five beverages with respect to the spici-
ness-reduction effect and hedonic impression with spicy chicken
meat. *** represents a significant difference between beverages
(P<0.001). Mean ratings with different small letters indicate a sig-
nificant difference between the beverages at P<0.05. Error bars rep-
resent standard error of the means.
Table 1 Pearson’s correlation coefficients in the relationship of self-reported rating for spiciness-reduction effect with the time-intensity curve
parameters.
Time-intensity curve parameters
Time at the
maximum
intensity
Maximum
intensity Duration
Area
under
the curve
Increasing
area
Increasing
angle
Decreasing
area
Decreasing
angle
Initial
delay
Initial
intensity
0.27*** 0.34*** 0.31*** 0.56*** 0.42*** 0.20** 0.45*** 0.03 0.04 0.09
**and ***indicate a significance at P<0.01 and P<0.001, respectively.
©2016 Institute of Food Science and TechnologyInternational Journal of Food Science and Technology 2016
Palate cleanser for spicy chicken meat S. S. Samant et al.2412
spiciness intensity (Duizer et al., 1995; Reinbach et al.,
2009). Therefore, beverages having lower values of the
TI parameters would be favoured choices in accor-
dance with the aim of this study, that is, spiciness-
reduction effect. The T
max
is an indication of how
quickly the participants reached maximum intensity.
Higher values indicate that the maximum spiciness
intensity was perceived slowly whereas lower values
indicate quicker perception. In the present study, T
max
was not significantly different among the five beverage
samples. However, the I
max
, AUC and Dec area, as
indicators of spiciness intensity, significantly differed
among the beverage samples. Milk (3.3% fat) had the
lowest values for I
max
, AUC and Dec area, thereby
making it the best beverage choice to reduce residual
spiciness elicited by spicy chicken consumption. It is
worth noting that the Dec area parameter in this case
exclusively focuses on spiciness reduction as it is the
area of the declining portion of the TI curve. More
specifically, lower values of Dec area for milk indicate
less residual spiciness perception (Duizer et al., 1995).
Effectiveness of milk on reducing burning sensations
of capsaicin in the mouth has been reported in the
past (Nasrawi & Pangborn, 1990; Lee & Kim, 2013).
It is worth pondering whether the fat content in milk
is responsible for reducing spiciness intensity as fats
have shown potential to be good palate cleansers
(Hutchinson et al., 1990). The effect of different fat
content in milk as a palate cleanser was studied by
Nasrawi and Pangborn (1990). They compared whole
milk and skim milk as potential palate cleanser for
spicy food. It was accomplished that even though both
types of milk were able to reduce the spiciness, whole
milk was slightly better than skim milk. Another study
by Lee & Kim (2013) studied milk-based palate cleans-
ers with varying fat and sucrose content and evaluated
their ability to reduce burning sensation due to cap-
saicin on a 15-point intensity scale. In their study, high
fat content was found to cause maximum reduction in
burning sensation. Therefore, milk might be the best
beverage option to reduce spiciness perception in the
mouth for consumers who do not care about the fat
content. However, there is a class of consumers who
are more calorie-conscious and prefer a healthy alter-
native. For these consumers, tomato juice could be
another possible beverage option. From the present
study, ability of tomato juice to reduce spiciness, in
terms of TI parameters, was slightly better than other
beverage samples, oolong tea, sparkling water and
spring water (Fig. 5).
Carbonation characteristic of beverage was found to
show no significant influence on the spiciness-reduction
effect in this study. Previous research has demon-
strated that both carbonation and spicy foods tend to
cause burning or stinging sensations on the tongue,
which are perceived via trigeminal nerve receptors
(Cartsens et al., 2002). However, our findings demon-
strate no significant difference between sparkling water
and spring water with respect to not only the TI
parameters, but also the self-reported rating of spici-
ness-reduction effect.
It was found that oolong tea is not superior in miti-
gating residual spiciness elicited by spicy food con-
sumption. In addition, participants did not appreciate
oolong tea not only as a spiciness reducer, but also as
a pair with spicy chicken. One possible reason for low
hedonic impression is the novelty of drinking tea with
spicy food. Moreover, the tea was served at room tem-
perature, but not at higher temperature at which it is
generally consumed. Therefore, even though there was
some finding of oolong tea being paired with spicy
food (Dornenburg & Page, 2009), it needs to be fur-
ther explored whether the oolong tea is suitable for
palate cleanser or pairing beverage with spicy chicken
products.
This study determined the relationship between self-
reported spiciness-reduction effect and a more objec-
tive estimate of TI parameters. Irrespective of the bev-
erage, six TI parameters, such as time at the maximum
intensity, maximum intensity, area under the curve,
duration, increasing area and decreasing area, could
well correspond to participants’ individual self-ratings
of spiciness-reduction effect. In particular, three
parameters, the area under the curve (PLSR
b=0.18; Pearson’s r=0.56), decreasing area
(PLSR b=0.14; Pearson’s r=0.45), and increas-
ing area (PLSR b=0.13; Pearson’s r=0.42), were
found to play as a major role in predicting the self-
reported rating of spiciness-reduction effect. Thus,
these TI parameters can be used for determining the
best option of palate cleanser for spicy foods.
The present findings also suggest possible food-bev-
erage pairings to ensure wholesome enjoyment of spicy
food experience. As shown in Fig. 6, even though
spring water could not well decrease residual spiciness
elicited by spicy chicken consumption, participants
rated spring water as a favourable beverage when they
consume spicy chicken. This result indicates while
some consumers prefer beverage mitigating residual
spiciness, other consumers prefer beverage lasting the
residual spiciness; this is in line with the notion that
some consumers have a high desirability for spicy food
(‘likers’) while other consumers almost averse it (‘non-
likers’) (Tepper et al., 2004).
Finally, it is worth noting that individual panellists’
variation in spiciness perception is evident in many
spiciness-related studies, including the present one.
This is understandable because, as mentioned earlier,
perception of spiciness intensity is dependent on the
concepts of sensitisation and desensitisation. It is diffi-
cult to generalise capsaicin sensitisation and desensiti-
sation due to high individual variation, which is also
©2016 Institute of Food Science and Technology International Journal of Food Science and Technology 2016
Palate cleanser for spicy chicken meat S. S. Samant et al. 2413
governed by genetic factors (Prescott, 1999; Prescott &
Stevenson, 1995). These phenomena affect the linger-
ing impact of spicy food on individuals. In other
words, some people perceive lingering residual spici-
ness for a longer period of time after spicy food con-
sumption compared to others. However, comparing
different maximum durations of residual spiciness was
beyond the scope of this study and can be proposed as
a future perspective. This information can also help in
optimising the time intervals between consecutive spicy
food presentations during sensory analysis to reduce
carry-over effects. Another interesting futuristic
approach could be verifying the current findings of TI
analysis with other existing and novel descriptive anal-
ysis methods (Dos Santos et al., 2015). In addition, as
the TI analysis is useful in finding an optimal forma-
tion (Rodrigues et al., 2015; De Morais et al., 2013),
the findings of this study could be extended to optimal
formulation of new spicy food or beverage products.
Furthermore, information from TI analysis, along with
consumer acceptance test, can help in development
and marketing of novel beverages intended to be
consumed with spicy food to reduce oral burn.
Conclusion
The present study demonstrates that milk (3.3% fat) is
more effective in reducing residual spiciness elicited by
spicy chicken consumption when compared to oolong
tea, tomato juice, sparkling water and spring water.
Thus, milk (3.3% fat) can be used as a palate cleanser
during sensory analysis of multiple spicy chicken prod-
ucts. In addition, the three time-intensity curve param-
eters, area under the curve, decreasing area and
increasing area, were found to best predict the self-
reported ratings for spiciness-reduction effect of indi-
vidual beverage samples. Finally, this study shows that
some consumers prefer beverage reducing residual
spiciness, whereas others also prefer beverage
lasting the residual spiciness elicited by spicy food
consumption.
Conflict of interest
The authors declare no conflict of interest.
References
Byrnes, N.K. & Hayes, H.E. (2015). Gender differences in the influ-
ence of personality traits on spicy food liking and intake. Food
Quality and Preference,42,12–19.
Cadena, R.S. & Bolini, H.M.A. (2011). Time-intensity analysis and
acceptance test for traditional and light vanilla ice cream. Food
Research International,44, 677–683.
Cadena, R.S., Cruz, A.G., Faria, J.A.F. & Bolini, H.M.A. (2012).
Reduced fat and sugar vanilla ice creams: sensor profiling and
external preference mapping. Journal of Dairy Science,95, 4842–
4850.
Cadena, R.S., Cruz, A.G., Netto, R.R., Castro, W.F., Faria, J.A.F.
& Bolini, H.M.A. (2013). Sensory profile and physiochemical
characteristics of mango nectar sweetened with high intensity
sweeteners through storage time. Food Research International,54,
1670–1679.
Cartsens, E., Carstens, M.I., Dessirier, J.M., O’Mahony, M.,
Simons, C.T. & Sudo, M. (2002). It hurts so good: oral irritation
by spices and carbonated drinks and the underlying neural mecha-
nisms. Food Quality and Preference,13, 431–443.
Cartwright, L.C., Snell, C.T. & Kelley, P.H. (1952). Organoleptic
panel testing as a research tool. Analytical Chemistry,24,
503–506.
Cliff, M. & Heymann, H. (1993). Time-intensity evaluation of oral
burn. Journal of Sensory Studies,8, 201–211.
Cox, I. & Gaudard, M. (2013). A deeper understanding of PLS. In:
Discovering Partial Least Squares with JMP. (edited by I. Cox &
M. Gaudard) Pp. 37–75 Cary: SAS Institute Inc..
De Morais, E.C., Cruz, A.G. & Bolini, H.M.A. (2013). Gluten-free
bread: multiple time-intensity analysis, physical, characterization
and acceptance test. International Journal of Food Science and
Technology,48, 2176–2184.
Dornenburg, A. & Page, K. (2009). What to Drink with What You
Eat: the Definitive Guide to Pairing Food with Wine, Beer, Spirits,
Coffee, Tea –Even Water –Based on Expert Advice from America’s
Best Sommeliers (1st ed., eBook). P. 84105. New York, USA: Bul-
finch Press.
Dos Santos, B.A., Campagnol, P.C.B., da Cruz, A.G. et al. (2015).
Check all that apply and free listing to describe the sensory charac-
teristics of low sodium dry fermented sausages: comparison with
trained panel. Food Research International,76, 725–734.
Duizer, L.M., Bloom, K. & Findlay, C.J. (1995). The effects of line
orientation on the recording of time-intensity perception of sweet-
ener solutions. Food Quality and Preference,6, 121–126.
Green, B.G. (1986). Sensory interactions between capsaicin and tem-
perature in the oral cavity. Chemical Senses,11, 371–382.
Hutchinson, S.E., Trantow, L.A. & Vickers, Z.M. (1990). The effec-
tiveness of common foods for reduction of capsaicin burn. Journal
of Sensory Studies,4, 157–164.
Kitsawad, K. & Nguyen, T.T. (2014). Determination of suitable
palate cleanser for spicy tom yum soup. Agro Food Industry Hi
Tech,25,28–30.
Lawless, H. T. & Heymann, H. (2010). Sensory Evaluation of Food –
Principles and Practices. Pp. 179–201. New York, USA: Springer.
Lawless, H., Rozin, P. & Shenker, J. (1985). Effects of oral capsaicin
on gustatory, olfactory, and irritant sensations and flavor identifi-
cation in humans who regularly or rarely consume chili pepper.
Chemical Senses,10, 579–589.
Lee, K.W. & Kim, K.O. (2013). Effects of fat and sucrose in palate
cleansers on discrimination of burning sensation of capsaicin sam-
ples. Food Science and Biotechnology,22, 691–696.
Lucak, C. & Delwiche, J.F. (2009). Efficacy of various palate cleans-
ers with representative foods. Chemosensory Perception,2,
32–39.
Luckett, C.R., Meullenet, J.F. & Seo, H.S. (2016). Crispness level of
potato chips affects temporal dynamics of flavor perception and
mastication patterns in adults of different age groups. Food Quality
and Preference,51,8–19.
Ludy, M.J. & Mattes, R.D. (2012). Comparison of sensory, physio-
logical, personality, and cultural attributes in regular spicy food
users and non-users. Appetite,58,19–27.
Meilgaard, M. C., Civille, G. V. & Carr, B. T. (2007). Sensory Evalu-
ation Techniques (4th ed.). Pp. 173–193. Boca Raton, USA: CRC
Press.
Nasrawi, C.W. & Pangborn, R.M. (1990). Temporal effectiveness of
mouth rinsing on capsaicin burn. Physiology and Behavior,47,
617–623.
©2016 Institute of Food Science and TechnologyInternational Journal of Food Science and Technology 2016
Palate cleanser for spicy chicken meat S. S. Samant et al.2414
Palazzo, A.B. & Bolini, H.M.A. (2009). Multiple time-intensity anal-
ysis and acceptance of raspberry-flavored gelatin. Journal of Sen-
sory Studies,24, 648–663.
Peyvieux, C. & Dijksterhius, G. (2001). Training a sensory panel for
TI: a case study. Food Quality and Preference,12,19–28.
Prescott, J. (1999). The generalizability of capsaicin sensitization and
desensitization. Physiology and Behavior,66, 741–749.
Prescott, J. & Stevenson, R.J. (1995). Pungency in food perception
and preference. Food Reviews International,11, 665–698.
Reinbach, H.C., Toft, M. & Moller, P. (2009). Relationship between
oral burn and temperature in chilli spiced pork patties evaluated
by time-intensity. Food Quality and Preference,20,
42–49.
Rodrigues, J.B., Paix~
ao, J.A., Cruz, A.G. & Bolini, H.M.A. (2015).
Chocolate milk with chia oil: ideal sweetness, sweeteners equiva-
lence, and dynamic sensory evaluation using a time-intensity
methodology. Journal of Food Science,80, S2944–S2949.
Rossini, K., Verdun, S., Cariou, V., Qannari, E.M. & Fogliatto, FS.
(2012). PLS discriminant analysis applied to conventional sensory
profiling data. Food Quality and Preference,23,18–24.
Simmons, C.T., O’Mahony, M. & Cartsens, E. (2002). Taste sup-
pression following lingual capsaicin pre-treatment in humans.
Chemical Senses,27, 353–365.
Stevens, D.A. & Lawless, H.T. (1986). Putting out the fire: effects of
tastants on oral chemical irritation. Perception and Psychophysics,
39, 346–350.
Tepper, B.J., Keller, K.L. & Ullrich, N.V. (2004). Genetic variation
in taste and preferences for bitter and pungent foods: implications
for chronic disease risk. In: Challenges in Taste Chemistry and Biol-
ogy. (edited by T.F. Hoffmann, C.T. Ho & W. Pickenhagen) Pp.
60–74 Washington, DC, USA: ACS Publications.
Wold, S., Sj€
ostr€
om, M. & Eriksson, L. (2001). PLS-regression: a
basic tool of chemometrics. Chemometrics and Intelligent Labora-
tory Systems,58, 109–130.
©2016 Institute of Food Science and Technology International Journal of Food Science and Technology 2016
Palate cleanser for spicy chicken meat S. S. Samant et al. 2415
