Odontology (2014) 102:6875
DOI 10.1007/s10266-012-0089-7
ORIGINAL ARTICLE
A two-colored chewing gum test for assessing masticatory
performance: a preliminary study
Toshiya Endo Akira Komatsuzaki
Hiroomi Kurokawa Satoshi Tanaka
Yoshiki Kobayashi Koji Kojima
Received: 5 April 2012 / Accepted: 1 October 2012 / Published online: 18 October 2012
The Society of The Nippon Dental University 2012
Abstract This study was conducted to compare sub-
jective and objective assessment methods of a two-colored
chewing gum test and to find out whether these methods
are capable of discriminating masticatory performances
between sexes. 31 adults, 16 males and 15 females par-
ticipated in this study. Each subject chewed five samples of
two-colored chewing gum sticks for 5, 10, 20, 30 and 50
chewing strokes, respectively. The subjective color-mixing
and shape indices for the gum bolus (SCMI-B, SSI-B) and
the subjective color-mixing index and objective color-
mixing ratio for the gum wafer (SCMI-W, OCMR-W) were
evaluated by two independent examiners and, on a different
day, re-evaluated by one of the examiners. The SCMI-B
and SCMI-W assessments had inter- and intra-examiner
reliable agreement at 20 or more chewing strokes. The
T. Endo (&)
Y. Kobayashi
Orthodontic Dentistry, The Nippon Dental University Niigata
Hospital, 1-8 Hamaura-cho, Chuo-ku, Niigata 951-8580, Japan
e-mail: endoto@ngt.ndu.ac.jp
A. Komatsuzaki
Department of Preventive and Community Dentistry,
The Nippon Dental University School of Life Dentistry
at Niigata, 1-8 Hamaura-cho, Chuo-ku, Niigata 951-8580, Japan
H. Kurokawa
Comprehensive Dentistry, The Nippon Dental University Niigata
Hospital, 1-8 Hamaura-cho, Chuo-ku, Niigata 951-8580, Japan
S. Tanaka
Department of Pediatric Dentistry, The Nippon Dental
University School of Life Dentistry at Niigata,
1-8 Hamaura-cho, Chuo-ku, Niigata 951-8580, Japan
K. Kojima
Koji Orthodontics Kids Dental Clinic, 3-2-11 Sakae,
Ichinomiya, Aichi 491-0858, Japan
123
OCMR-W measurement demonstrated high accuracy and
low reproducibility between and within the examiners.
There were significant gender differences in the distribu-
tion of SCMI-W scores (P = 0.044) and in the mean
OCMI-W (P = 0.007). The SCMI-B and SCMI-W
assessments and the OCMR-W measurement were reliable
and valid at the 20 and 30 chewing strokes in this two-
colored chewing gum test. The subjective color-mixing
index (SCMI-W) and objective color-mixing ratio (OCMR-W)
for the chewing gum wafer are capable of discriminating
masticatory performance between sexes in this two-
colored chewing gum test and that the OCMR-W mea-
surement is discriminating better than the SCMI-W
assessment.
Keywords Two-colored chewing gum
Masticatory
performance
Color-mixing index
Bolus-shaping index

Color-mixing ratio
Introduction
Mastication is a complex process characterized by the
comminution and breakdown of food into small particles to
increase its surface area, stimulate an enzymatic function
and improve the digestibility of food [1]. Masticatory
efficiency can be measured by counting the number of
chewing strokes required to reduce food to a certain par-
ticle size before swallowing, and masticatory performance
is measured by assessing the particle size distribution of
food when chewed for a given number of strokes [2].
A food comminution test with a sieving method has
been typically used to test masticatory performance [36].
However, it is said that the comminution test is too com-
plicated to apply clinically and needs too much time for
Odontology (2014) 102:6875
testing and evaluating [7]. One of the reasons for the use of
the comminution test may be that the main function of
mastication is to comminute food. Some foods need to be
comminuted while others need to flow or form a bolus
before swallowing. The physical properties of foods should
be considered before the study of masticatory performance
is conducted [8].
Masticatory performance is influenced by several factors,
including age [4, 5, 9], sex [4, 5, 9], body size [4, 5], number
and type of teeth in occlusion [3, 9, 10], neuro-physiological
deficits [11] and type and severity of malocclusions [5, 6, 9].
The association between malocclusion and masticatory
performance is of clinical importance because orthodontic
treatment is made to achieve correct occlusion in harmony
with form and function. Several studies have reported the
influence of surgical orthodontic treatments on masticatory
performance [12, 13]. Others have reported the negative
effects of malocclusion on masticatory performance [5, 6, 9].
Therefore, it is important to evaluate the quantification of
masticatory performance in clinical orthodontics so that it
becomes possible to determine how orthodontic treatments
influence oral function.
Two-colored chewing gum [10, 11, 1418] and paraffin
wax [7, 1921] have been used as test foods for assessing
the degree of bolus formation and color-mixing and
quantifying masticatory performance. Color-mixing of the
chewed test food has been assessed objectively by com-
puter analysis of digital images, [7, 1820] subjectively by
visual inspection [10, 15, 16] or by both [14, 17, 21].
Before assessing the mixing ability with the two-colored
chewing gum and wax by computer analysis of the digital
images, the chewed bolus must be flattered to a wafer to
avoid high contrast shallows in the image by the oblique
illumination of the scanners lamp [17, 18, 21]. Flattening
of the chewed bolus provides a more accurate and reliable
assessment of masticatory performance than observation
of the bolus in their raw state [14, 17]. Some pieces of
evidence demonstrate that the mixing ability test with the
two-colored chewing gum or wax is a valid and reliable
alternative for food comminution tests [18, 19]. The mixing
ability test with the two-colored chewing gum is a good
method to determine masticatory function in subjects with
a compromised masticatory performance [18]. Computer
analysis of the digital images of chewed two-colored wax
is more suitable for discriminating between the mixing
abilities of groups of participants with a limited mastica-
tory performance than visual assessment of these images
[21]. Some investigators report that visual assessment of
the chewed bolus has similar accuracy as computer
assessment of digital image [17]. Others show that visual
assessment is less reliable than computer assessment, but it
may still useful in screening of chewing deficiencies in a
clinical setting [14].
69
Some researchers have reported no sex differences in
masticatory performance using the food comminution tests
[5, 9]. Others have showed the mean or median particle
size for males significantly smaller than for females [4, 22].
The possibility of sex difference in masticatory perfor-
mance still remains controversial.
The purpose of this preliminary study was to compare
subjective and objective assessment methods of a two-
colored chewing gum test and to find out whether these
methods are capable of discriminating masticatory perfor-
mance between males and females.
Materials and methods
Test subjects
The protocol was approved by the Research Ethics Com-
mittee of the Nippon Dental University at Niigata.
Informed consent was obtained from all participants in this
study, who had been given written and/or oral information
about this project.
Subjects comprised 31 volunteers, 16 males and 15
females, who were either dental students or staff members
of The Nippon Dental University Niigata Hospital (Niigata,
Japan). The selection criteria of the subjects were (1) a
complete permanent dentition with fully erupted incisors to
second molar at each quadrant, (2) all premolars and
molars with at least one occlusal contact in habitual
occlusion, (3) no signs of temporomandibular joint (TMJ)
dysfunction, (4) Class I occlusion, (5) little or no incisor
crowding, (6) no history of previous orthodontic or pros-
thetic treatment, (7) no mesiodistal or class II restorations,
crowns or onlays and (8) no tooth anomalies. Mean ages of
males and females were 24.8 years (SD 2.3 years) and
23.8 years (SD 1.5 years), respectively, with no statisti-
cally significant difference between sexes (P = 0.192).
Chewing gum
A commercially available two-colored chewing gum used
for the assessment of masticatory performance was stained
blue and red by the manufacturer (Lotte, Tokyo, Japan;
distributor, Oral Care, Tokyo, Japan) with water-insoluble
dyes. A test piece that measured 18.8 9 14.2 9 3.9 mm
was made from one blue and one red piece.
Chewing test procedure
All the subjects were examined and instructed by one
operator (K.K.). They were asked to sit in a dental chair
with their heads held in a natural position. Each subject
was told to chew five samples of two-colored chewing gum
123
Odontology (2014) 102:6875
testing and evaluating [7]. One of the reasons for the use of
the comminution test may be that the main function of
mastication is to comminute food. Some foods need to be
comminuted while others need to flow or form a bolus
before swallowing. The physical properties of foods should
be considered before the study of masticatory performance
is conducted [8].
Masticatory performance is influenced by several factors,
including age [4, 5, 9], sex [4, 5, 9], body size [4, 5], number
and type of teeth in occlusion [3, 9, 10], neuro-physiological
deficits [11] and type and severity of malocclusions [5, 6, 9].
The association between malocclusion and masticatory
performance is of clinical importance because orthodontic
treatment is made to achieve correct occlusion in harmony
with form and function. Several studies have reported the
influence of surgical orthodontic treatments on masticatory
performance [12, 13]. Others have reported the negative
effects of malocclusion on masticatory performance [5, 6, 9].
Therefore, it is important to evaluate the quantification of
masticatory performance in clinical orthodontics so that it
becomes possible to determine how orthodontic treatments
influence oral function.
Two-colored chewing gum [10, 11, 1418] and paraffin
wax [7, 1921] have been used as test foods for assessing
the degree of bolus formation and color-mixing and
quantifying masticatory performance. Color-mixing of the
chewed test food has been assessed objectively by com-
puter analysis of digital images, [7, 1820] subjectively by
visual inspection [10, 15, 16] or by both [14, 17, 21].
Before assessing the mixing ability with the two-colored
chewing gum and wax by computer analysis of the digital
images, the chewed bolus must be flattered to a wafer to
avoid high contrast shallows in the image by the oblique
illumination of the scanners lamp [17, 18, 21]. Flattening
of the chewed bolus provides a more accurate and reliable
assessment of masticatory performance than observation
of the bolus in their raw state [14, 17]. Some pieces of
evidence demonstrate that the mixing ability test with the
two-colored chewing gum or wax is a valid and reliable
alternative for food comminution tests [18, 19]. The mixing
ability test with the two-colored chewing gum is a good
method to determine masticatory function in subjects with
a compromised masticatory performance [18]. Computer
analysis of the digital images of chewed two-colored wax
is more suitable for discriminating between the mixing
abilities of groups of participants with a limited mastica-
tory performance than visual assessment of these images
[21]. Some investigators report that visual assessment of
the chewed bolus has similar accuracy as computer
assessment of digital image [17]. Others show that visual
assessment is less reliable than computer assessment, but it
may still useful in screening of chewing deficiencies in a
clinical setting [14].
69
Some researchers have reported no sex differences in
masticatory performance using the food comminution tests
[5, 9]. Others have showed the mean or median particle
size for males significantly smaller than for females [4, 22].
The possibility of sex difference in masticatory perfor-
mance still remains controversial.
The purpose of this preliminary study was to compare
subjective and objective assessment methods of a two-
colored chewing gum test and to find out whether these
methods are capable of discriminating masticatory perfor-
mance between males and females.
Materials and methods
Test subjects
The protocol was approved by the Research Ethics Com-
mittee of the Nippon Dental University at Niigata.
Informed consent was obtained from all participants in this
study, who had been given written and/or oral information
about this project.
Subjects comprised 31 volunteers, 16 males and 15
females, who were either dental students or staff members
of The Nippon Dental University Niigata Hospital (Niigata,
Japan). The selection criteria of the subjects were (1) a
complete permanent dentition with fully erupted incisors to
second molar at each quadrant, (2) all premolars and
molars with at least one occlusal contact in habitual
occlusion, (3) no signs of temporomandibular joint (TMJ)
dysfunction, (4) Class I occlusion, (5) little or no incisor
crowding, (6) no history of previous orthodontic or pros-
thetic treatment, (7) no mesiodistal or class II restorations,
crowns or onlays and (8) no tooth anomalies. Mean ages of
males and females were 24.8 years (SD 2.3 years) and
23.8 years (SD 1.5 years), respectively, with no statisti-
cally significant difference between sexes (P = 0.192).
Chewing gum
A commercially available two-colored chewing gum used
for the assessment of masticatory performance was stained
blue and red by the manufacturer (Lotte, Tokyo, Japan;
distributor, Oral Care, Tokyo, Japan) with water-insoluble
dyes. A test piece that measured 18.8 9 14.2 9 3.9 mm
was made from one blue and one red piece.
Chewing test procedure
All the subjects were examined and instructed by one
operator (K.K.). They were asked to sit in a dental chair
with their heads held in a natural position. Each subject
was told to chew five samples of two-colored chewing gum
123
70
Fig. 1 The subjective color-mixing index for the chewing gum bolus
(SCMI-B). Scores 1 through 5
on their preferred chewing side for 5, 10, 20, 30 and 50
chewing strokes, respectively, and to terminate the act of
chewing with their mouth closed. Thereafter, the instructor
had the subjects open the mouth and removed the gum
bolus with an explorer. The number of chewing strokes was
counted by both subject and instructor. The subjects were
recommended to take a rest for at least 1 min between
trials.
Subjective visual assessments
The chewing gum bolus removed from the mouth after
each stroke test was coded and listed by a co-worker (S.T.).
The color-mixing and shape of the coded chewing gum
bolus were inspected visually and assessed subjectively by
an examiner (H.K.). To avoid any examination bias, the
assessment was made blind to the sex and the number of
chewing strokes.
The subjective color-mixing index for the chewing gum
bolus (SCMI-B) was determined in the way Schimmel
et al. [14] used in their experiment. As shown in Fig. 1, the
SCMI-B scores ranged from 1 to 5: 1, chewing gum not
mixed, impressions of cusps or folded once; 2, large parts
of chewing gum unmixed; 3, bolus slightly mixed, but bits
of unmixed original color; 4, bolus well mixed, but color
not uniform; 5, bolus perfectly mixed with uniform color.
The subjective shape index for the chewing gum bolus
(SSI-B) was determined in accordance with the method
used by Liedberg and Owall [16]. As shown in Fig. 2, the
SSI-B scores ranged from 1 to 4: 1, flattened, square or
round; 2, ball; 3, long, spool-shaped, round cross-section;
4, long, spool-shaped, flattened.
To evaluate the inter-examiner agreement with respect
to the SCMI-B and SSI-B assessments, each coded bolus
was scored once again by another examiner (A.K.) 10 min
later. The determination of these scores was also made
blind to the sex of the subjects and the number of chewing
strokes.
After the determinations of SCMI-B and SSI-B scores,
the chewing gum bolus was flattened to a wafer of 1 mm
thickness with the same code as the gum bolus and the
123
Odontology (2014) 102:6875
Fig. 2 The subjective shape index for the chewing gum bolus (SSI-
B). Scores 1 through 4
subjective color-mixing index for the chewing gum wafer
(SCMI-W) was determined in the way Schimmel et al. [14]
used in their experiment. The chewed gum was rinsed and
placed between two sheets of stiff and clear plastics. The
sandwich of plastic and gum was squashed flat between
two thick plates with a 1 mm spacer between them to
provide a uniformly thick wafer. The SCMI-W assessment
was made independently by the same examiners (H.K.,
A.K.) using the same scores and blinding design as the
SCMI-B assessment.
Objective computerized measurement
The objective color-mixing ratio for the coded chewing
gum wafer (OCMR-W) was determined with computerized
measurements the same way as was done by Schimmel
et al. [3] in their study. Each wafer was scanned from either
side of the wafer with a resolution of 500 dots per inch
using a high-quality scanner (ES-2200, Seiko Epson,
Nagano, Japan). The scanned image was copied into an
image of fixed size (1227 9 1227 pixels) and stored in
Adobe Photoshop CS3 format (Adobe, San Jose, CA,
USA). Magnetic Lasso Tool was used to select the whole
area of the chewing gum wafer on both sides and then the
numbers of pixels were measured from the histogram panel
for this area. As a reference scale, a scanned image of the
unmixed blue gum was copied (74 9 165 pixels) and
stored in the same manner with the wafer image. Then the
color range was used (tolerance 100) to select the unmixed
blue parts of the wafer image. The numbers of selected
pixels were measured from the histogram panel for each
side of the unmixed blue parts of the wafer image.
Thereafter, the OCMR was calculated for the unmixed
fraction using the following formula: sum of the numbers
of pixels for the unmixed blue parts on the wafer surface
scanned from both sides/sum of the numbers of pixels in
the entire area of the double-faced wafer images 9 100
Odontology (2014) 102:6875
(%). The OCMR-W was measured by the two independent
examiners (H.K., A.K.) blindly to the sex and the number
of chewing strokes.
To evaluate the intra-examiner agreement for the SCMI-
B, SSI-B and SCMI-W assessments and the intra-examiner
accuracy and reproducibility of the OCMR-W measure-
ment, all the subjects were called back to repeat the same
experiment on a different day. To get rid of the stress and
fatigue, all subjects were asked to sit in the dental chair and
take a rest for at least 10 min before the chewing tests after
confirmation that no dental and medical treatments were
received from the first examination. These assessments and
measurement were made for the daytime by the same
examiner (H.K.) blind to the sex of the subjects and the
number of chewing strokes. Afterward the results of the
examinations were sorted by these parameters for statistical
comparisons.
Statistical analysis
Statistical analyses were performed by the use of SPSS
Mac version 17.0J (SPSS Japan, Tokyo, Japan) and Excel-
Toukei for Windows (SSRI, Tokyo, Japan). Weighted
kappa values and standard errors were used to test the inter-
and intra-examiner agreements for the SCMI-B, SSI-B and
SCMI-W assessments after each chewing test. The
weighted kappa value represents the degree of agreement
such as: \0 corresponding to poor agreement; 0 to 0.20,
slight agreement; 021 to 0.40, fair agreement; 0.41 to 0.60,
moderate agreement; 061 to 0.80, substantial agreement;
081 to 1.00, almost perfect agreement [23]. To evaluate the
inter- and intra-examiner accuracy and reproducibility of
the OCMR-W measurements after each stroke test in each
sex, the systematic error was determined by a paired t test
with a 95 % confidence interval and the random error was
determined by calculating the standard deviation of the
differences of two measurements [24].
Normality of the OCMR-W for all chewing strokes was
tested with the skewness-kurtosis test. A two-way analysis
of variance (ANOVA) with GreenhouseGeisser correction
and Bonferroni post hoc test was performed to test the main
effects of the numbers of chewing strokes and the sexes on
Table 1 Inter-examiner agreement (kappa value) for the SCMI-B, SSI-B and SCMI-W assessments
Number of chewing strokes
5 10
j value SE j value SE
SCMI-B 0.52 0.16
SSI-B 0.07
SCMI-W 0.23 0.21
SE standard error
71
the distribution of subjects by the SCMI-B, SSI-B and
SCMI-W scores and on mean OCMR-W. The two-way
ANOVA showed a significant interaction between two
variables in the SCMI-B and SSI-B assessments, Friedman
and Bonferroni post hoc tests, and Pearson Chi-square test
were used to determine whether and/or where there were
significant differences in the distribution of subjects
between the numbers of chewing strokes in each sex and
between sexes in each number of chewing strokes,
respectively.
Sample size was not performed, because the current
study was a pilot study. However, previous researches on
mixing ability tests demonstrated significant differences
with sample sizes C20 [3, 5, 6, 16, 18, 20, 21].
Results
As shown in Tables 1 and 2, all chewing gum specimens
which were chewed for 5 strokes in the SCMI-B, SSI-B
and SCMI-W assessments and for 10 strokes in the SSI-B
assessment were classified into category 1 by the two
examiners, thus inter- and intra-examiner kappa values
could not be calculated because of the lack of variation.
Slight or fair inter-examiner agreement was found in the
SSI-B assessments at 20, 30 and 50 chewing strokes and
the SCMI-W assessment at 10 chewing strokes, and mod-
erate agreement occurred in the SCMI-B and SCMI-W
assessments at other chewing strokes (Table 1). Except for
the kappa values (fair intra-examiner agreement) of the
SCMI-B and SCMI-W assessments at 10 chewing strokes
(0.27 and 0.28, respectively), the overall kappa values were
calculated at 0.44 or more, indicating moderate to almost
perfect intra-examiner agreement (Table 2). In the SCMI-
B, SSI-B and SCMI-W assessments, all chewing gum
specimens which were chewed for 5 strokes were classified
into category 1, thus verifying that the two-colored
chewing gum at five chewing strokes was not valid for
discriminating masticatory performance. On the basis of
the inter- and intra-examiner agreement and the distribu-
tion of scores, the SSI-B assessment at all chewing strokes
and the SCMI-B and SCMI-W assessments at 5 and 10
20 30 50
j value SE j value SE j value SE
0.50 0.13 0.42 0.15 0.51 0.10
0.07 0.12 0.07 0.10 0.07
0.45 0.15 0.59 0.13 0.46 0.11
123
72 Odontology (2014) 102:6875

Table 2 Intra-examiner agreement (kappa value) for the SCMI-B, SSI-B and SCMI-W assessments
Number of chewing strokes
5 10 20 30 50
j value SE j value SE j value SE j value SE j value SE

SCMI-B 0.27 0.18 0.65 0.14 0.90 0.07 0.78 0.10

SSI-B 0.44 0.17 0.49 0.15 0.52 0.17
SCMI-W 0.28 0.22 0.54 0.15 0.57 0.13 0.47 0.13
SE standard error

Table 3 Two-way ANOVA

Source Sum of squares df Mean square F value P value
for the distribution of the SCMI-
B, SSI-B, SCMI-W scores and SCMI-B
the mean OCMR-W as a
function of number of chewing Number of chewing strokes 233.174 3.052 76.396 362.414 0.000***
strokes and sex Interaction 2.206 3.052 0.723 3.429 0.020*
Error 18.658 88.514 0.211
Sex 0.871 1 0.871 1.559 0.222NS
SSI-B
Number of chewing strokes 154.105 3.148 48.959 329.900 0.000***
Interaction 1.924 3.148 0.611 4.119 0.008**
Error 13.547 91.281 0.148
Sex 2.999 1 2.999 11.854 0.002**
SCMI-W
Number of chewing strokes 249.862 4 62.465 415.441 0.000***
Interaction 1.216 4 0.304 2.022 0.096NS
Error 17.442 116 0.150
Sex 1.680 1 1.680 4.423 0.044*
OCMR-W
Number of chewing strokes 29664.786 2.638 11245.912 180.604 0.000***
Interaction 322.483 2.638 122.253 1.963 0.134NS
NS not significant Error 4763.341 76.497 62.268
* P \ 0.05, ** P \ 0.01, Sex 603.331 1 603.331 8.392 0.007**
*** P \ 0.001

chewing strokes were not reliable or valid and conse-
quently left out of consideration.
Table 3 shows the results of the two-way ANOVA with
Greenhouse-Geisser correction. A significant difference
was observed in the distribution of SCMI-B scores among
the numbers of chewing strokes (P = 0.000) but no sig-
nificant difference was noted between sexes (P = 0.222).
A significant interaction was found between these two
variables (P = 0.020). Pearson Chi-square tests demon-
strated no significant differences in the distribution of
SCMI-B scores at upward of 20 chewing strokes between
sexes, and the Friedman and Bonferroni post hoc tests
demonstrated significant differences between 20 and 30
chewing strokes, between 20 and 50 chewing strokes, and
between 30 and 50 chewing strokes in both sexes.
From Table 3, it is evident that the two-way ANOVA found
significant differences in the distribution of SCMI-W scores
among the numbers of chewing strokes (P = 0.000) and
between sexes (P = 0.044). However, no significant interac-
tion was noted between these two variables (P = 0.096).
As shown in Tables 4 and 5, no systematic errors in the
OCMR-W measurement between or within the examiners
were found at 20 chewing strokes and over irrespective of
sex. For males and females, the random errors in the
OCMR-W measurement between the examiners were 1.2
and 3.5 %, 0.5 and 0.8 % and 0.1 and 0.2 % at 20, 30 and
50 chewing strokes, respectively (Table 4). Likewise, the
random errors for males and females in the OCMR-W
measurement within the examiners were 1.8 and 4.2 %, 0.3
and 0.9 %, and 0.1 and 0.5 % at 20, 30 and 50 chewing
strokes, respectively (Table 5).
The skewnesskurtosis test showed a normal distribution
of the OCMR-W (P = 0.3512). The mean OCMR-W and
standard deviation at each number of chewing strokes in each

123
Odontology (2014) 102:6875 73

Table 4 Systematic and random errors in the OCMR-W measure- sex are presented in Table 6. As shown in Table 3, the two-
ment between the examiners way ANOVA with Greenhouse-Geisser correction demon-
Number of chewing strokes strated significant differences in the mean OCMI-W among
the numbers of chewing strokes (P = 0.000) and between
5 10 20 30 50
sexes (P = 0.007) and no significant interaction between
Male these two variables (P = 0.134). Table 3 also shows that
Systematic 0.383NS 0.676NS 0.115NS 0.729NS 0.948NS SCMI-W and OCMR-W are both capable to discriminate
error (P value) between males and females and that OCMR-W is discrimi-
Random error 6.8 2.9 1.2 0.5 0.1 nating better (P = 0.007) than SCMI-W (P = 0.044).
(%)
Table 6 shows that the Bonferroni post hoc test demonstrates
Female significant differences in mean OCMR-W between 20 and 30
Systematic 0.101NS 0.626NS 0.998NS 0.190NS 0.433NS chewing strokes and between 20 and 50 chewing strokes but
error (P value)
not between 30 and 50 chewing strokes.
Random error 5.1 4.1 3.5 0.8 0.2
(%)
NS not significant
Discussion

In this study, the SCMI-B and SCMI-W assessments at 20,

Table 5 Systematic and random errors in the OCMR-W measure- 30 and 50 chewing strokes showed moderate inter-examiner
ment within the examiners agreement and moderate to almost perfect intra-examiner
agreement. These results suggest that the subjective assess-
Number of chewing strokes
ments of the samples run on the two-color chewing gum test
5 10 20 30 50 generates the inter- and intra-examiner reliable agreement at
Male 20 chewing strokes and longer.
Systematic 0.058NS 0.624NS 0.769NS 0.640NS 0.680NS Our results also showed that the inter- and intra-exam-
error (P value) iner kappa values for the SCMI-B and SCMI-W assess-
Random error 4.8 2.9 1.8 0.3 0.1 ments at 20 and 30 chewing strokes ranged between 0.42
(%) and 0.59 and between 0.54 and 0.9, respectively. Investi-
Female gating different methods of assessing the degree of color
Systematic 0.075NS 0.784NS 0.864NS 0.438NS 0.732NS mixture of the bolus and wafer using a kind of commer-
error (P value) cially available two-color chewing gum (Hubba-Bubba
Random error 5.4 3.9 4.2 0.9 0.5 Tape Gums), Schimmel et al. [14] found that the inter- and
(%)
intra-examiner kappa values for the gums chewed for 20
NS not significant and 30 strokes ranged between 0.47 and 0.66 and between

Table 6 The OCMR-W and statistical comparisons

Number of chewing strokes
5 (5S) 10 (10S) 20 (20S) 30 (30S) 50 (50S)
Mean SD Mean SD Mean SD Mean SD Mean SD

Males
33.8 9.5 17.1 10.7 3.1 3.9 0.5 0.7 0.1 0.1
Females
38.8 9.2 24.5 9.8 9.3 8.0 1.5 2.8 0.3 0.5
Statistical comparison between the numbers of chewing strokes/Bonferroni test/P value
5S vs. 10S 0.000***
5S vs. 20S 0.000*** 10S vs. 20S 0.000***
5S vs 30S 0.000*** 10S vs 30S 0.000*** 20S vs 30S 0.003**
5S vs. 50S 0.000*** 10S vs. 50S 0.000*** 20S vs 50S 0.000*** 30S vs 50S 0.367 NS
SD standard deviation, NS not significant
** P \ 0.01, *** P \ 0.001

123
74
0.2 and 0.78, respectively. van der Bilt et al. [21] showed
that the inter- and intra-examiner kappa values for visual
assessment of the two-colored wax chewed for 520
strokes varied between 0.7 and 0.78 and between 0.55 and
0.65, respectively. High inter- and intra-examiner kappa
values in our study, as well as their studies, demonstrated
that the two-color chewing gum test would serve the pur-
pose of assessing masticatory performance. The SSI-B
assessment was not relevant because of slight inter-exam-
iner agreement in this study, whereas Liedberg et al. [16]
showed that there were no significant differences in both
color-mixing and bolus-shape indices between the exam-
iners using the commercially available two-color SOR-
BITS chewing gums and that bolus shaping as well as
color-mixing was relevant as the chewing gum parameter.
These different results may be due to some difference in
quality between two-color chewing gums used and suggest
that the inter- and intra-examiner agreement needs to be
investigated, when other brands of chewing gum are used
for clinical as well as research purposes.
Our results showed no significant systematic errors in
the OCMR-W measurement between or within the exam-
iners at 20 chewing strokes and longer, thus showing that
this measurement had high inter- and intra-examiner
accuracy. In this study, the mean of the OCMR-W was
higher than, or equal to the random error in the OCMR-W
measurement between and within the examiners at each of
20 chewing strokes and longer. Moreover, the shorter the
number of chewing strokes was, the larger the discrepancy
between the mean of the OCMR-W and the random error
was, thus proving that the reproducibility of OCMR-W
measurements was higher at 20 chewing strokes and lower
at 50 chewing strokes in the 2050 chewing stroke tests.
These findings on the random errors may suggest that the
20- and 30-stroke chewing sequences could be used for
the OCMR-W measurement. However, when the large
standard deviations of the OCMR-W are taken into consid-
eration, it must be said that the random errors in the OCMR-
W measurement indicate low inter- and intra-examiner
reproducibility. The computerized OCMR-W measurement
with the commercially available software package including
scanning, pixel counting and subsequent mathematical cal-
culations are not realistic in a clinical setting, as stated by
Schimmel et al. [14]. The development of software and
automated instrument for the OCMR-W measurement and
the training of examiners may be needed to reduce random
errors and to improve measurement reproducibility.
Recently, van der Bilt et al. [21] showed that computer
analysis of digital images of the chewed two-colored wax
was better than visual assessment of these images in dis-
criminating the mixing of colors in the wax when it was not
very well mixed. Thus, computer method can be very good
in the quantification of masticatory performance. Our
123
Odontology (2014) 102:6875
results support these findings in that the SCMI-W assess-
ment and the OCMR-W measurement are capable of dis-
criminating masticatory performance between sexes and
that the OCMR-W measurement is discriminating better
than the SCMI-W assessment, which can be seen from
P values in Table 3.
In this study, there were significant differences between
the sexes in the distribution of SCMI-W scores and in mean
OCMR-W, as confirmed by the two-way ANOVA. These
results were supported by Julien et al. [4], who estimated
the median particle size using the seven-sieve method and
showed significant differences in masticatory performance
between men (2235 years of age) and women
(2229 years of age). However, Shiere and Manly [9]
showed no significant difference in masticatory perfor-
mance between boys and girls at age 79, and 1214. A
probable reason for different results may be the difference
in subject ages. Julien et al. [4] and we selected adult
subjects, while Shiere and Manly [9] used young boys and
girls as their subjects. In the large cross sectional study,
Shiere and Manly [9] reported that the masticatory per-
formance increased between 6 and 10 years of age,
decreased between 10 and 11 years of age, and then
increased again by age 14, thus suggesting that the age was
one of the most important factors in masticatory perfor-
mance, as evidenced by Julien et al. [4], and Toro et al. [5].
van den Braber et al. [12, 13] adopted a silicone rubber
Optosoft and 15- and 30-stroke chewing sequences to
investigate the effects of surgical orthodontic treatment on
masticatory performance in adult patients (mean age
24.8 years) with mandibular retrognathism. Shiere and
Manly [9] used peanuts and either 60 or 80 chewing strokes
to evaluate the change of chewing ability with age in
children of 615 years. Toro et al. [5] and English et al. [6]
used the condensation silicone impression material, Cut-
terSil and a 20-stroke chewing sequence to investigate the
association of malocclusion types and masticatory perfor-
mance in children (5.515.5 years of age) and adolescents
(737 years of age). van der Bilt et al. [18] evaluated the
mixing abilities of two groups of young subjects with a
natural dentition (mean age 24 years) and elderly subjects
mostly with complete dentures (mean age 72 years) using
the two-colored chewing gum with 520 chewing strokes
and Optosil and Optocal with 15 chewing strokes, and
found that the mixing ability test with this chewing gum
was a good method to determine masticatory function in
subjects with a compromised masticatory performance
(elderly subjects). In using the two-colored chewing wax
and 520 chewing strokes to discriminating between the
masticatory performances of three groups of subjects with
a natural dentition, full dentures, and maxillary denture
plus implant-supported mandibular overdenture, van der
Bilt et al. [21] showed that computer analysis of digital
Odontology (2014) 102:6875
images of the chewed two-colored wax could discriminate
the mixing abilities of three groups and visual assessment
of these images could distinguish the mixing abilities of the
dentate subjects from the two denture wearer groups but
not between the two denture groups. The discrepancy in the
proper number of chewing strokes may be attributed to
differences in each combination of test food (peanuts,
Optosil, chewing gum, chewing wax, and so on) and group
of participants (young dentates, elderly edentates, dental
implants, orthodontic patients, and so on).
In this study, the gums which were chewed for 20, 30
and 50 strokes were consistently classified into two or three
categories of the SCMI-B and SCMI-W scores. This vari-
ation of categories will make it possible to differentiate
between subjects. Moreover, significant differences in the
distribution of SCMI-B and SCMI-W scores were found
between 20 and 30 chewing strokes, and between 30 and 50
chewing strokes, and there were significant differences in
mean OCMR-W between 20 and 30 chewing strokes but
not between 30 and 50 chewing strokes. These results
suggest that the 20- or 30-stroke chewing sequence is
suitable for assessing masticatory performance using this
two-colored chewing gum test.
In conclusion, the subjective color-mixing index (SCMI-
W) and objective color-mixing ratio (OCMR-W) for the
chewing gum wafer are capable of discriminating masti-
catory performance between sexes in this two-colored
chewing gum test and that the OCMR-W measurement is
discriminating better than the SCMI-W assessment.
Conflict of interest The authors declare that they have no conflict
of interest.
References
1. Magalhaes IB, Pereira LJ, Marques LS, Gameiro GH. The
influence of malocclusion on masticatory performance: a sys-
tematic review. Angle Orthod. 2010;80:9817.
2. Bates JF, Stafford GD, Harrison A. Masticatory functiona
review of the literature. III. Masticatory performance and effi-
ciency. J Oral Rehabil. 1976;3:5767.
3. Van der Bilt A, Olthoff LW, Oosterhaven SP. The effect of
missing postcanine teeth on chewing performance in man. Arch
Oral Biol. 1993;38:4239.
4. Julien KC, Buschang PH, Throckmorton GS, Dechow PC. Nor-
mal masticatory performance in young adults and children. Arch
Oral Biol. 1996;41:6975.
5. Toro A, Buschang PH, Throckmorton G, Roldan S. Masticatory
performance in children and adolescents with Class I and II
malocclusions. Eur J Orthod. 2006;28:1129.
75
6. English JD, Buschang PH, Throckmorton GS. Does malocclusion
affect masticatory performance? Angle Orthod. 2002;72:217.
7. Sato H, Fueki K, Sueda S, Sato S, Shiozaki T, Kato M, Ohyama
T. A new and simple method for evaluating masticatory function
using newly developed artificial test food. J Oral Rehabil.
2003;30:6873.
8. Lucas PW, Luke DA. Is food particle size a criterion for the
initiation of swallowing? J Oral Rehabil. 1986;13:12736.
9. Shiere FR, Manly RS. The effect of the changing dentition on
masticatory function. J Dent Res. 1952;31:52634.
10. Liedberg B, Spiechowicz E, Owall B. Mastication with and
without removable partial dentures: an intraindividual study.
Dysphagia. 1995;10:10712.
11. Schimmel M, Leemann B, Herrmann FR, Kiliaridis S, Schnider
A, Muller F. Masticatory function and bite force in stroke
patients. J Dent Res. 2011;90:2304.
12. van den Braber W, van der Glas H, van der Bilt A, Bosman F.
Masticatory function in retrognathic patients, before and after
mandibular advancement surgery. J Oral Maxillofac Surg.
2004;62:54954.
13. van den Braber W, van der Bilt A, van der Glas H, Rosenberg T,
Koole R. The Influence of mandibular advancement surgery on
oral function in retrognathic patients: a 5-year follow-up study.
J Oral Maxillofac Surg. 2006;64:123740.
14. Schimmel M, Christou P, Herrmann F, Muller F. A two-colour
chewing gum test for masticatory efficiency: development of
different assessment methods. J Oral Rehabil. 2007;34:6718.
15. Liedberg B, Owall B. Masticatory ability in experimentally
induced xerostomia. Dysphagia. 1991;6:2113.
16. Liedberg B, Owall B. Oral bolus kneading and shaping measured
with chewing gum. Dysphagia. 1995;10:1016.
17. Prinz JF. Quantitative evaluation of the effect of bolus size and
number of chewing strokes on the intra-oral mixing of a two-
colour chewing gum. J Oral Rehabil. 1999;26:2437.
18. van der Bilt A, Mojet J, Tekamp FA, Abbink JH. Comparing
masticatory performance and mixing ability. J Oral Rehabil.
2010;37:7984.
19. Sato S, Fueki K, Sato H, Sueda S, Shiozaki T, Kato M, Ohyama
T. Validity and reliability of a newly developed method for
evaluating masticatory function using discriminant analysis.
J Oral Rehabil. 2003;30:14651.
20. Speksnijder CM, Abbink JH, van der Glas HW, Janssen NG, van
der Bilt A. Mixing ability test compared with a comminution test
in persons with normal and compromised masticatory perfor-
mance. Eur J Oral Sci. 2009;117:5806.
21. van der Bilt A, Speksnijder CM, de Liz Pocztaruk R. Digital
image processing versus visual assessment of chewed two-colour
wax in mixing ability tests. J Oral Rehabil. 2012;39:117.
22. Wilding RJC, Lewin A. The determination of optimal human jaw
movements based on their association with chewing performance.
Arch oral Biol. 1994;39:33343.
23. Landis RJ, Koch GG. The measurement of observer agreement
for categorical data. Biometrics. 1977;33:15974.
24. Othman S, Harradine N. Tooth size discrepancies in an ortho-
dontic population. Angle Orthod. 2007;77:66874.
123