ORIGINAL ARTICLE

Comparison of intermaxillary tooth size discrepancies among

different malocclusion groups

Qiong Nie, DDS, MS, PhD,a and Jiuxiang Lin, DDS, MS, PhDb
Beijing, China

The purpose of this study was to determine whether there is a prevalent tendency for intermaxillary tooth
size discrepancies among different malocclusion groups. This study consisted of 60 subjects who served as
the normal occlusion group and 300 patients divided into 5 malocclusion groups (ie, Class I with bimaxillary
protrusion, Class II Division 1, Class II Division 2, Class III, and Class III surgery). Tooth size measurements
were performed on the models of normal occlusion and pretreatment models of patients by the Three
Dimension Measuring Machine. Moreover, tooth size ratios, analyzed as described by Bolton and the
Student t test showed no sexual dimorphism for these ratios in each of 6 groups, so the sexes were
combined for each group. Then these ratios were compared among different malocclusion groups. The
results showed no significant difference between subcategories of malocclusion, so these groups were
combined. There were now 120 cases in each of 3 categories: Class I, Class II, and Class III. A significant
difference was found for all the ratios between the groups, the ratios showing that Class III > Class I > Class
II. It demonstrated that intermaxillary tooth size discrepancy may be one of the important factors in the
cause of malocclusions, especially in Class II and Class III malocclusions. Thus this study proved the fact
that Bolton analysis should be taken into consideration during orthodontic diagnosis and therapy. (Am J
Orthod Dentofacial Orthop 1999;116:539-44)

Bolton, in 1958, developed a method of

1 tooth size between sexes, as reported by a number of
analyzing mesiodistal tooth size ratio between maxil-
lary and mandibular teeth. In his study, he concluded
that it would be difficult for proper occlusal interdigita-
tion or coordination of arches in the finishing stage of
orthodontic treatment without proper mesiodistal tooth
size ratio between maxillary and mandibular teeth.
Stifer2 replicated Bolton study in Class I dentitions
and reported similar results. Subsequently, other
authors obtained the normal values of Bolton analysis
of different races, eg, that of a Chinese population.3
These studies were performed on a mixture of
treated and untreated subjects with good or excellent
occlusion. However, up to now, for the study of tooth
size of malocclusion, especially for the comparison of
intermaxillary tooth size relationship among different
malocclusions, few studies were available and the
research results were controversial.
Arya et al4 showed that there were differences in
aAssistant Professor, Department of Orthodontics, School of Stomatology, Bei-
jing Medical University.
bProfessor, Department of Orthodontics, School of Stomatology, Beijing Med-
other authors. They attempted to show differences in
tooth size between Class I and Class II malocclusions,
but failed to do so. In their study, the mean size of each
tooth for the different groups (ie, Class I and Class II,
boys and girls) were compared. Difference for individ-
uals between different arches were not analyzed.
Lavelle5 showed that there was sexual dimorphism
in tooth dimensions and in the ratio of upper to lower
arch tooth size. In addition, there was racial dimor-
phism between Negroids, Mongoloids, and Cau-
casians. Lavelle also measured the ratio of upper to
lower arch tooth size in different malocclusion types,
as in the current study. The difference was in the
method of analysis. Rather than tooth sizes being
compared for individuals, the mean size of each tooth
of male patients for each malocclusion type was
stated. A pattern of contrast was found, which was
different for the maxillary values, compared with
mandibular values for different malocclusion cate-
gories. In this study, the mesiodistal crown dimen-
sions for maxillary teeth were Class I > Class II Divi-
sion 1 > Class II Division 2 > Class III, cf, Class III >

ical University; Vice President, Beijing Medical University.
Reprint requests to: Dr Qiong Nie, Department of Orthodontics, School of
Stomatology, Beijing Medical University, No 38, BaiShiQiao Road, HaiDian
District, Beijing, 100081 China; e-mail, Liuyq@infoc3.icas.ac.cn
Copyright © 1999 by the American Association of Orthodontists.
0889-5406/99/$8.00 + 0 8/1/100624
Class I > Class II Division 1 > Class II Division 2 for
mandibular teeth. It can be inferred that as a general
trend, the Bolton discrepancy would be greater in
Class III cases than other malocclusion groups but
this was not analyzed for individuals.
539
540 Nie and Lin
Fig 1. YM-2115 Three Dimension Measuring Machine.
The Sperry6 study analyzed the Bolton ratios for
groups of Class I, Class II, and Class III cases. The
skeletal patterns were not mentioned, although some of
the Class III cases were treated surgically. Male and
female subjects were not differentiated. The overall
ratios showed that there was mandibular tooth-size
excess for the Class III patients.
Crosby and Alexander7 analyzed the Bolton ratios
for different occlusal categories. They did not differen-
tiate between sexes, and they did not include Class III
patients. The relationship of malocclusion to skeletal
pattern was not mentioned. They did not find a statisti-
cally significant difference in the prevalence of tooth
size discrepancies among the different malocclusion
groups.
The objectives of the current study were to deter-
mine (1) whether sexual dimorphism exists for tooth
size ratios, and (2) whether there is a difference for
intermaxillary tooth size discrepancies represented by
anterior ratio, overall, and posterior ratio of Bolton, for
Class I, Class II Division 1, Class II Division 2, Class
III, and Class III surgery cases.
MATERIAL AND METHODS
The samples for the study consisted of 60 subjects
with normal occlusion and 300 patients with varying
American Journal of Orthodontics and Dentofacial Orthopedics
November 1999
malocclusions. Patients were selected randomly from
clinical practice of the department of orthodontics,
School of Stomatology, Beijing Medical University in
the 1990s. Subjects with normal occlusion were
selected from the study of Growth and Development in
this department (project financially supported by
National Nature and Science Foundation of China)
whose models were recorded in the 1990s. All cases
were Han nationality born and living in China. They
were between 13 and 17 years old except Class III
surgery patients who were from 17 to 23 years old.
Occlusion categories of all cases, which were classified
by the Angle classification, coincided with skeletal cat-
egories. Skeletal types were assessed by ANB angle
from cephalometric analysis, which meant in skeletal
Class I, ANB angle was from 0o to 5o, cf, ANB angle >
5o for Skeletal Class II and ANB angle < –2o for skele-
tal Class III. All subjects were divided into 6 groups;
each group consisted of 30 males and 30 females.
These groups were normal occlusion, Class I with
bimaxillary protrusion, Class II Division 1, Class II
Division 2, Class III, and Class III surgery cases. The
selection criteria of a Class III surgery case was severe
skeletal Class III malocclusion mainly represented by
severe mandibular prognathism and an ANB angle <
–7o. In this study, the models were the major investiga-
tion focus.
The following selection criteria were used:
1. Good quality models of normal occlusion and pretreat-
ment models of patients.
2. All permanent teeth had erupted and were present from
right first molar through left first molar.
3. No severe mesiodistal and occlusal tooth abrasion.
4. No residual crown or crown-bridge restoration.
5. No tooth deformity (eg, conic-form lateral incisal teeth)
Each model was oriented in a YM-2115 Three
Dimension Measuring Machine (similar to Yamada et
al8 and Braun et al9) as shown in Fig 1. This sort of
machine is used extensively in the precision machine
tool industry. In this study, the device was a miniature
system. It ran in the range of 150 mm × 200 mm × 100
mm and the accuracy of the 3 orthogonal axes was all
0.01 mm. The frictionless air bearing and touch trigger
probe was used to identify the measuring point (ie,
anatomic contact point of each tooth), and to record the
corresponding X, Y, and Z coordinates automatically to
a computer data file. Through the calculation of a com-
puter program, the greatest mesiodistal diameters of all
the teeth on each cast were obtained except the second
and third molars. In order to determine the measure-
ment error, 5 sets of models were measured again sev-
eral days later by the same examiner. The result
showed no significant difference between the 2 mea-
American Journal of Orthodontics and Dentofacial Orthopedics Nie and Lin 541
Volume 116, Number 5

Table I. Tooth size ratios for 6 groups

Female patients Male patients

χ SE SD χ SE SD

Normal occlusion
RA 81.10 .60 2.27 81.95 .42 2.28
RP 104.24 .48 2.63 104.19 .59 3.23
RO 93.11 .48 2.64 93.44 .43 2.35
Bimaxillary protrusion
RA 81.25 .52 2.87 81.87 .46 2.51
RP 104.75 .66 3.61 104.61 .59 3.24
RO 93.41 .46 2.53 93.62 .44 2.42
Class II Division 1
RA 80.80 .44 2.42 80.31 .71 3.87
RP 102.74 .70 3.86 103.35 .61 3.34
RO 92.21 .44 2.39 92.11 .48 2.61
Class II Division 2
RA 80.97 .49 2.66 81.07 .64 3.52
RP 101.91 .60 3.28 102.41 .53 2.92
RO 91.82 .41 2.26 92.09 .49 2.70
Class III
RA 83.10 .61 3.33 82.60 .57 3.12
RP 107.14 .57 3.12 107.60 .90 4.92
RO 95.62 .44 2.43 95.49 .55 3.01
Class III surgery
RA 82.67 .46 2.51 82.61 .50 2.75
RP 107.31 .58 3.17 108.36 .63 3.47
RO 95.41 .45 2.44 95.86 .46 2.52
Malocclusion
RA 81.76 .24 2.90 81.69 .27 3.27
RP 104.77 .33 4.03 105.26 .35 4.29
RO 93.69 .23 2.86 93.83 .25 3.08

surements. Then tooth size ratios were analyzed as RESULTS

Bolton described: Table I summarizes the means, standard deviations,
and standard error of the tooth size ratios observed in
Sum mandibular 6 – 6
}}} × 100 = overall ratio (RO) each group. It shows that there is no significant sexual
Sum maxillary 6 – 6
dimorphism for all ratios of all groups. As to the
absolute value of tooth size ratios of different sexes,
Sum mandibular 3 – 3
}}} × 100 = anterior ratio (RA) there is no prevalent trend. For instance, anterior ratio
Sum maxillary 3 – 3
of male patients (81.07 ± 3.52) was larger than that of
female patients (80.97 ± 2.66) in Class II Division 2,
Sum mandibular 654 – 456
}}} × 100 = posterior ratio (RP) but that of male patients (80.31 ± 3.87) was smaller
Sum maxillary 654 – 456
than that of female patients (80.80 ± 2.42) in Class II
Division 1.
In order to determine whether there is sexual ANOVA demonstrated that there were significant
dimorphism in the incidence of intermaxillary tooth differences among the 6 groups (tables omitted, avail-
size discrepancies, a Student t test was performed for able from authors), thus Student-Newman-Keuls Mul-
each malocclusion group. Moreover, in order to ticomparison was performed to analyze the differences.
compare intermaxillary tooth size discrepancies Because there was no significant sexual dimor-
among different malocclusion groups, ANOVA and phism for tooth size ratios, the sexes were combined
Student-Newman-Keuls Multicomparisons were per- for each group. Multicomparison was then performed
formed. The software of above statistical analysis between these groups. The results are shown in Table
was SPSS (version 6.0), and the level of significance II. It was obvious that for the 3 tooth size ratios, there
was P < .05. are no significant differences between Class III and
542 Nie and Lin American Journal of Orthodontics and Dentofacial Orthopedics
November 1999

Table II. Comparison of tooth size ratios among the 6 groups

χ SD Group 1 2 3 4 5 6 Subset and order

RA
Normal occlusion 81.52 2.82 1 1:Grp 1,2,3,4
Bimaxillary protrusion 81.56 2.69 2 2:Grp 1,2,5,6
Class II Division 1 80.56 3.21 3 Grp 3<4<1<2<6<5
Class II Division 2 81.02 3.10 4
Class III 82.85 3.20 5 * *
Class III surgery 82.64 2.61 6 * *
RP
Normal occlusion 104.21 2.92 1 1:Grp3,4
Bimaxillary protrusion 104.68 3.40 2 * 2:Grp 1,3
Class II Division 1 103.03 3.59 3 * 3:Grp1,2
Class II Division 2 102.16 3.09 4 * 4:Grp 5,6
Class III 107.37 4.09 5 * * * * Grp 4<3<1<2<5<6
Class III surgery 107.84 3.34 6 * * * *
RO
Normal occlusion 93.27 2.48 1 1:Grp 3,4
Bimaxillary protrusion 93.51 2.46 2 2:Grp 1,2
Class II Division 1 92.16 2.48 3 * * 3:Grp 5,6
Class II Division 2 91.95 2.47 4 * * Grp 4<3<1<2<5<6
Class III 95.55 2.71 5 * * * *
Class III surgery 95.63 2.47 6 * * * *

*Level of significance of Multicomparison is P < .05.

Subset: there were no significant differences among the groups.

Table III. Comparison of tooth size ratios among the three occlusion categories
χ SD Group 1 2 3 Subset and order

RA
Class I 0.8154 0.0275 1 Grp 2<1<3
Class II 0.8079 0.0315 2 *
Class III 0.8275 0.0291 3 * *
RP
Class I 1.0445 0.0317 1 Grp 2<1<3
Class II 1.0260 0.0336 2 *
Class III 1.0760 0.0373 3 * *
RO
Class I 0.9339 0.0246 1 Grp 2<1<3
Class II 0.9206 0.0247 2 *
Class III 0.9559 0.0258 3 * *

*Level of significance of Multicomparison is P < .05.

Class III surgery, between normal occlusion and Class
I bimaxillary protrusion, between Class II Division 1
and Class II Division 2.
Because there is no significant difference between
subcategories of malocclusion, these groups are com-
bined. There are now 120 cases for each category of
Angle classification. Then the Multicomparison was
performed between 3 new groups and the statistical
results were summarized in Table III. It shows that the
tendencies of anterior ratio, posterior ratio, and overall
ratio were all Class III > Class I > Class II, the differ-
ences between these groups were significant at the
level of significance P < .05. For example, overall
ratios of the 3 categories were 95.59 ± 2.58, 93.39 ±
2.46, 92.06 ± 2.47, respectively.
DISCUSSION
When the whole malocclusion samples (ie, 150
female or 150 male patients) were combined into 1
group, our study shows that tooth size ratios of the
malocclusion group are close to that of the normal
occlusion group, as shown in Table I. For example, the
overall ratio of the normal occlusion group for female
patients is 93.11 ± 2.64; for the malocclusion group it
American Journal of Orthodontics and Dentofacial Orthopedics
Volume 116, Number 5
was 93.69 ± 2.86. Zhu Xia and Xiying Wu10 also
found no significant difference for tooth size ratios
between the malocclusion group and the normal occlu-
sion group after measuring mesiodistal tooth sizes of
1173 Han nationality cases on their models. Thus, only
after comparing tooth size ratios among different clas-
sified malocclusion groups, can the law of nature be
observed.
Sperry et al6 showed that the Class III group with
mandibular prognathism had more patients with
mandibular tooth-size excess for the overall ratio than
the Class I and Class II groups (0.01 < P < .05). This
conclusion was similar to a result of the present study.
The similar result was that the overall ratio of Class III
surgery was the highest among different malocclusion
groups. However, the present study demonstrated that
not only Class III surgery but also Class III nonsurgery
had a greater frequency of mandibular tooth size excess
than other malocclusion groups.
Lavelle5 showed that tooth sizes of Class III were
the smallest among the 3 occlusion categories (ie, Class
I, Class II and Class III) for maxillary teeth; they were
the greatest for mandibular teeth. This possibly indi-
cated that tooth size ratios of mandibular teeth divided
by maxillary teeth in Class III may be the greatest
among different malocclusion types. However, these
ratios were not compared in his study. His result was
only a kind of descriptive statistical result, which stated
the mean size of each tooth of male patients for each
malocclusion type and described a pattern of contrast.
The present study compared these ratios and showed
that anterior ratio, posterior ratio, and overall ratio of
Class III malocclusions were all greater than other
occlusion categories.
Crosby and Alexander7 also compared the tooth
size ratios among different malocclusion groups, as in
the current study. They found that there were no sig-
nificant differences among Class I, Class II Division 1,
Class II Division 2, and Class II surgery groups. The
current study also determined that no significant dif-
ferences exist between Class II Division 1 and Class II
Division 2, but other results were different from their
study. The present study showed that the 3 tooth size
ratios were all Class III > Class I > Class II and the dif-
ferences between these groups were significant. The
differences of results between the current study and
Crosby and Alexander 7 study could be explained as
follows.
First of all, the skeletal categories were not men-
tioned in Crosby and Alexander’s study although some
of Class II cases were treated surgically. This may have
an important effect on the selection of sample. For
example, some skeletal Class II malocclusions can be
Nie and Lin 543
converted to dental Class I malocclusions by forward
movement of permanent first molar due to the prema-
ture loss of the deciduous second molar, so the Class I
group may contain skeletal Class I and Class II
patients. In the current study, skeletal categories were
taken into account, and in order to simplify this study,
the cases were selected by the criteria of occlusal cate-
gories coinciding with skeletal categories.
Second, Crosby and Alexander7 did not differenti-
ate between sexes and did not mention the ratio of
sexes in each group. In their study, it was not clear
whether there was sexual dimorphism for tooth size
ratios. The present study separated sexes and demon-
strated that there was no sexual dimorphism for tooth
size ratios, thus the sexes were combined in the ratio of
1:1 for each group.
Third, Crosby and Alexander7 did not include Class
III patients in their study and only selected 20 to 30
cases for each group. However, in the current study, not
only Class III but also Class III surgery patients were
included and 60 cases were contained in each of the 6
groups. As mentioned previously, because there was no
significant difference between subcategories of maloc-
clusion, these groups could be combined. There were
120 cases in each category of Angle classification in
the present study. Therefore, the samples of the current
study were greater and classification of malocclusion
was more complete than their study.
Finally, the samples of the Class I group in the
Crosby and Alexander7 study were composed of Class
I malocclusion in which no prevalent clinical presenta-
tions were mentioned, but that of the present study
were made up of normal occlusion and Class I maloc-
clusion with bimaxillary protrusion. This may be
another reason that led to the differences in results
between their study and ours.
In clinical practice, clinicians often note the dis-
crepancy of tooth size and skeletal size but seldom pay
attention to tooth size discrepancy between maxillary
and mandibular teeth. The present study showed the
tendency of mandibular tooth size excess in Angle
Class III malocclusion and the tendency of maxillary
tooth size excess in Angle Class II malocclusion. This
indicated that it might be reasonable for orthodontists
to do interproximal stripping or tooth extraction in the
mandibular dentition for Class III malocclusion and in
the maxillary dentition for Class II malocclusion.
These results suggested that the Bolton analysis is
important and should be considered when diagnosing,
planning, and predicting prognosis in clinical ortho-
dontics.
In the current study, the Class I malocclusion group
only consisted of bimaxillary protrusion because of the
544 Nie and Lin
consideration of our other studies. This was a defect of
sample selection. But it demonstrated that Class I cases
with bimaxillary protrusion had no prevalent incidence
of intermaxillary tooth size discrepancy. The samples of
other clinical presentations are needed to be added into
the Class I group to further determine the relationship of
Class I to Class II or Class I to Class III. For Class II and
Class III malocclusions, this study showed a prevalent
tendency of intermaxillary tooth size discrepancy.
Although tooth size and tooth size ratios described
by Bolton were different in different racial groups, and
the order was Negroids > Mongoloids > Caucasoids.8
However, there are little data in relation to the degree
and frequency of intermaxillary tooth size discrepancy
in different racial groups for the same malocclusion
category. This study demonstrated a statistical ten-
dency of intermaxillary tooth size discrepancies for dif-
ferent malocclusion groups in a Chinese population.
Certainly, further work is needed to explain the proba-
ble existing racial differences for intermaxillary tooth
size discrepancies in different malocclusion categories.
CONCLUSION
Mesiodistal tooth sizes were measured in a total of
360 sets of casts for normal occlusion, Class I bimaxil-
lary protrusion, Class II Division 1, Class II Division 2,
Class III, and Class III surgery. Tooth size ratios were
analyzed as described by Bolton and compared among
different malocclusion groups. The results showed:
1. Significant sexual dimorphism for the 3 tooth size ratios
did not exist.
American Journal of Orthodontics and Dentofacial Orthopedics
November 1999
2. When tooth size ratios were compared, there were no sig-
nificant differences between normal occlusion and Class I
bimaxillary protrusion, between Class II Division 1 and
Class II Division 2, between Class III and Class III
surgery patients.
3. The tendencies of the 3 tooth-size ratios for the 3 occlu-
sion categories were all Class III > Class I > Class II at the
level of significance P < .05. It suggests that the tooth size
discrepancy between maxillary and mandibular teeth may
be one of the important factors in the cause of malocclu-
sions, especially in Class II and Class III malocclusions.
In order to obtain optimal and stable treatment results, the
Bolton analysis should be taken into consideration when
diagnosing, planning, and predicting prognosis in clinical
orthodontics.
REFERENCES
1. Bolton WA. Disharmony in tooth size and its relation to the analysis and treatment of
malocclusion. Angle Orthod 1958;28:113-130.
2. Stifer J. A study of Pont, Howes Rees Neff, and Bolton analyses on Class I adult den-
titions. Angle Orthod 1958;28:215.
3. Lin IX, Fu MK. Analysis of maxillary tooth size versus mandibular tooth size and its
clinical application. J Mod Stomatol 1991;5:17-8.
4. Arya BS, Savara BS, Thomas D, et al. Relation of sex and occlusion to mesiodistal
tooth size. Am J Orthod 1974;66:479-86.
5. Lavelle CLB. Maxillary and mandibular tooth size in different racial groups and in
different occlusion categories. Am J Orthod 1972;6:29-37.
6. Sperry TP, Worms FW, Isaacson RJ, et al. Tooth-size discrepancy in mandibular prog-
nathism. Am J Orthod 1977;72:183-90.
7. Crosby DR, Alexander CG. The occurrence of tooth size discrepancies among differ-
ent malocclusion groups. Am J Orthod Dentofacial Orthop 1989;95:457-61.
8. Yamada A, Nagahara K, Yuasa S, et al. A study of reliability in three dimensional mea-
surement of a dental casts. J Japanese Orthod Soc 1987;46:18-31.
9. Braun S, Hnat WP, Fender DE. The form of the human dental arch. Angle Orthod
1998;68:29-36.
10. Xia Z, Wu XY. The application of dentocclusal measurement in malocclusion. Stom-
atology 1983;3:126-7.