ORIGINAL ARTICLE

Effects of premolar extractions on Bolton overall

ratios and tooth-size discrepancies in a Japanese
orthodontic population
Toshiya Endo,a Katuyuki Ishida,b Isao Shundo,b Kosuke Sakaeda,c and Shohachi Shimookad
Niigata and Kawasaki, Japan

Introduction: The purpose of this study was to investigate the effects of premolar extractions on the Bolton
overall ratios and overall tooth-size discrepancies in a Japanese orthodontic population. Methods: Mesiodis-
tal tooth widths were measured on 198 pretreatment dental casts of subjects with Class I, Class II, and Class III
malocclusions. The overall ratios and tooth-size discrepancies were determined before and after hypothetical
premolar extractions. Before and after extractions, the subjects were divided into small, normal, and large
overall ratio groups categorized by the Bolton standard deviation definition, and into small, normal, and large
correction groups by the actual amount of change calculated for tooth-size correction in millimeters. Extrac-
tions were performed in the following combinations: (1) all first premolars, (2) all second premolars, (3) maxillary
first and mandibular second premolars, and (4) maxillary second and mandibular first premolars. For statistical
evaluations, analysis of variance, Kruskal-Wallis, Friedman, Scheffé, Bonferroni, and Tukey tests were per-
formed. Results: The overall ratios decreased in every malocclusion group after extraction of any combination
of premolars. The decreases were significantly notable in combinations 2 and 4. Some subjects in the normal
overall ratio and maxillary and mandibular correction groups moved into the clinically significant tooth-size dis-
crepancy group after premolar extraction, and the reverse was also true. Conclusions: In formulating a treat-
ment plan involving premolar extractions, orthodontists should consider that the overall ratios might decrease,
and normal and clinically significant tooth-size discrepancies could change mutually after extractions. (Am J
Orthod Dentofacial Orthop 2010;137:508-14)

malocclusion types.5,6 Others found no significant dif-

B
olton’s tooth-size ratios, including overall and
anterior ratios, have been widely accepted as
an essential diagnostic tool in orthodontic prac-
tice since their publication.1,2 The Bolton anterior and
overall ratios were defined as the ratios of the mesiodis-
tal widths between the 6 anterior mandibular teeth and
the 6 anterior maxillary teeth, and the mesiodistal
widths between the 12 mandibular teeth and the 12 max-
illary teeth, respectively.1,2 Some evidence points to sex
and racial or ethnic differences in the tooth-size ra-
tios.3,4 Some investigators reported statistically signifi-
cant associations between tooth-size ratios and
a
Professor and chairman, Orthodontic Dentistry, Nippon Dental University
Niigata Hospital, Niigata, Japan.
b
Assistant professor, Orthodontic Dentistry, Nippon Dental University Niigata
Hospital, Niigata, Japan.
c
Private practice, Kawasaki, Japan.
d
Professor and chairman, Department of Pediatric Dentistry, Nippon Dental
University School of Life Dentistry at Niigata, Niigata, Japan.
The authors report no commercial, proprietary, or financial interest in the prod-
ucts or companies described in this article.
Reprint requests to: Toshiya Endo, Orthodontic Dentistry, Nippon Dental Uni-
versity Niigata Hospital, 1-8 Hamaura-cho, Niigata 951-8580 Japan; e-mail,
endoto@ngt.ndu.ac.jp.
Submitted, February 2008; revised and accepted, April 2008.
0889-5406/$36.00
Copyright Ó 2010 by the American Association of Orthodontists.
doi:10.1016/j.ajodo.2008.04.026
508
ferences in either anterior or overall ratios in subjects
with different malocclusion.7-9
A tooth-size discrepancy between the maxillary and
mandibular arches is found by analyzing tooth-size ra-
tios. Some studies found that ratios more than 2 SD
from Bolton’s mean of 91.3% indicated clinically sig-
nificant overall tooth-size discrepancies.5-12 In ortho-
dontic practice, the actual amount of discrepancy (in
millimeters) provides more useful information on the
required correction for clinically significant tooth-size
discrepancy than does the ratio as a percentage.9,13
Some investigators9,13,14 selected 1.5 mm as an appro-
priate threshold for clinical significance of discrepancy,
quoting Proffit and Ackerman15 that tooth-size discrep-
ancies less than 1.5 mm were rarely significant.
In orthodontic treatment, tooth extraction is often
necessary to achieve the best possible esthetic and func-
tional outcome for patients, and the extraction of 4 first
premolars is most common.16 The overall ratio and
tooth-size discrepancies are directly influenced by pre-
molar extraction. Nonetheless, only a few investigations
have been conducted about the applicability of the over-
all ratio as a criterion, shown by the shortage of litera-
ture.2,17 Bolton2 found that the mean overall ratio was
American Journal of Orthodontics and Dentofacial Orthopedics
Volume 137, Number 4
91.3% (SD, 1.91) in patients without a tooth-size dis-
crepancy, and, after extraction of 4 premolars, the pa-
tients had a mean overall ratio of 88% (SD, 1). Tong
et al17 stated that the overall ratios after extraction of
all premolar combinations were smaller than those be-
fore extraction, and, in some of the patients, normal
and large overall ratios changed into small and normal
overall ratios, respectively, after extraction of premo-
lars. The change of overall ratio groups was especially
noteworthy in combinations of all second premolars
and the maxillary second and mandibular first premo-
lars. Saatci and Yukay18 and Gaidyte and Baubiniene19
investigated tooth-size discrepancies created by pre-
molar extractions using the Bolton index, which is a pos-
itive value of either maxillary or mandibular corrections
required to give the Bolton mean overall ratio.
The purpose of this study was to investigate the ef-
fects of premolar extractions on the Bolton overall ratios
and tooth-size discrepancies in a Japanese orthodontic
population.
MATERIAL AND METHODS
A total of 198 Japanese subjects with various maloc-
clusions were selected retrospectively from a list of or-
thodontic patients who had received treatment in our
clinics at the Nippon Dental University Niigata Hospi-
tal, Niigata, Japan. They included patients with Class
I, Class II, and Class III malocclusions and met the cri-
teria for the dental casts as described below. The occlu-
sion category, according to Angle’s classifications,
coincided with the skeletal category. Skeletal types
were assessed cephalometrically by the mean ANB an-
gles (3.3 6 2.1 for males; 2.6 6 1.7 for females) 20:
Class I, from 1.2 to 5.4 for males and from 0.9 to 4.3
for females; Class II, .5.4 for males and .4.3 for
females; and Class III, \1.2 for males and \0.9 for
females. Each malocclusion group consisted of 33
male and 33 female subjects. The selection criteria of
the casts were (1) fully erupted permanent dentition
with only the third molars unerupted, (2) good-quality
pretreatment casts, (3) no tooth agenesis or extractions,
(4) no mesiodistal restorations or abrasion, and (5) no
tooth anomalies.
Digital calipers were used to measure the mesiodis-
tal widths from first molar to first molar to the nearest
0.01 mm on each cast. The mesiodistal width of each
tooth was measured at the greatest distance between
the contact points on the proximal surfaces. All mea-
surements were done by 1 investigator (I.K.). The over-
all ratios were calculated by using the method of
Bolton.1,2
Endo et al 509
Thirty pairs of dental casts were randomly selected
a month later, and the mesiodistal tooth widths were
again measured by the same investigator. The overall ra-
tios were calculated by the same method. A paired t test
showed no statistically significant differences between
the first and second measurements (P .0.05). Random
errors, assessed by calculating the standard deviation of
the differences between the first and second measure-
ments, were less than 0.72% for the overall ratios, and
less than 0.8 and 0.73 mm for the maxillary and mandib-
ular corrections, respectively; these were unlikely to
affect the significant results in this study.13
Because 2-way analysis of variance (ANOVA) indi-
cated no significant differences in overall ratios between
the sexes or malocclusion types, and no significant inter-
action between 2 variables, the values for the sexes were
combined for all other analyses, as shown in Tables I
and II.
In each malocclusion group, hypothetical tooth ex-
tractions were performed on each subject in the follow-
ing 4 combinations: (1) all first premolars, (2) all second
premolars, (3) maxillary first and mandibular second
premolars, and (4) maxillary second and mandibular first
premolars. The overall ratios were again calculated after
the hypothetical extractions in each malocclusion group.
Calculations were made before extractions to deter-
mine the distributions of subjects with overall tooth-
size discrepancies more than 2 SD from the Bolton
means (91.3% 6 1.91%) and more than 1.5 mm of
maxillary or mandibular correction required to give
the Bolton mean overall ratio, and all the subjects
were subdivided into 3 groups in each category: (1)
small overall ratio (\87.48%), (2) normal overall ratio
(87.48% to 95.12%), and (3) large overall ratio
(.95.12%). Both maxillary and mandibular correction
groups were (1) small correction (\–1.5 mm), (2) nor-
mal correction (–1.5 to 11.5 mm), and (3) large correc-
tion (.1.5 mm).
Similarly, to determine how many subjects moved
into other groups, the number of subjects in each group
was calculated again according to the Bolton mean
overall ratio (88%; SD, 1)2 and a 1.5-mm threshold
after all premolar extraction combinations.
Statistical analysis
Statistical analyses were performed with StatMate
software (ATMS, Tokyo, Japan). The means and standard
deviations of the overall ratios were calculated before and
after the 4 extraction combinations of premolars in each
malocclusion group. Two-way ANOVA was performed
to test the main effects of extractions and malocclusion
types on the overall ratio. One-way ANOVA and Scheffé
510 Endo et al American Journal of Orthodontics and Dentofacial Orthopedics
April 2010

Table I. Overall ratios (mean 6 SD) before and after extractions in each malocclusion group and statistical compar-
isons
Before extraction (BE) After extraction Scheffe´

Group Male Female Both sexes 4/4 5/5 4/5 5/4 Significance

Class I 91.18 6 2.27 91.01 6 2.17 91.10 6 2.20 89.74 6 2.41 88.89 6 2.11 90.04 6 2.37 88.61 6 2.24 BE . 5/5, BE . 5/4
malocclusion
Class II 91.48 6 1.91 91.28 6 1.87 91.38 6 1.88 89.91 6 2.00 89.00 6 1.73 90.07 6 1.84 88.86 6 1.93 BE . 4/4, BE . 5/5,
malocclusion BE . 4/5, BE . 5/4
Class III 91.27 6 1.56 91.85 6 2.16 91.56 6 1.89 90.27 6 1.98 89.24 6 1.89 90.47 6 1.99 89.05 6 1.96 BE . 5/5, BE . 5/4,
malocclusion 4/5 . 5/4

4/4, All first premolars; 5/5, all second premolars; 4/5, maxillary first and mandibular second premolars; 5/4, maxillary second and mandibular first
premolars.

Table II.Two-way ANOVA of overall ratios before Table III.Two-way ANOVA of overall ratios as a func-
extractions as a function of sex and malocclusion type tion of extraction and malocclusion type
Source Sum of squares df Mean square F value P value Source Sum of squares df Mean square F value P value

Sex 0.231 1 0.231 0.057 0.811 Extraction 799.355 4 199.839 549.901 0.000
Malocclusion 7.154 2 3.577 0.890 0.412 Malocclusion 33.101 2 16.551 0.857 0.426
type type
Interaction 6.492 2 3.246 0.808 0.447 Interaction 2.183 8 0.273 0.751 0.646
Error 771.615 192 4.019 Error 283.459 780 0.363

tests were used to compare the overall ratios before and
after extractions in each malocclusion group and to iden-
tify where differences occurred. Kruskal-Wallis and
Bonferroni tests, and Friedman and Tukey tests, were per-
formed to determine whether and where there were
changes in the numbers of subjects and the distributions
of overall ratios, respectively, in each overall ratio group
and in each maxillary and mandibular correction group
related to the extraction combinations. ANOVA and mul-
tiple comparison tests were performed at the P \0.05
level of significance.
RESULTS
Table III shows that 2-way ANOVA indicated no
significant interaction between the main effects of ex-
tractions and malocclusion types and no significant dif-
ferences in the overall ratios between the malocclusion
types. It also shows significant difference in the overall
ratios between before and after extractions. Table I
shows that the overall ratios decreased in every maloc-
clusion group after extraction of any combination of
premolars, and that the statistically significant de-
creases were particularly notable in combinations 2
and 4.
Two-way ANOVA, performed to test the main ef-
fects of sex and malocclusion types on the overall ratios,
indicated no significant differences between malocclu-
sion types (Table II). Kruskal-Wallis tests showed no
significant differences between the malocclusion
groups in the distributions of subjects in the normal
overall ratio group and those of the clinically significant
overall tooth-size discrepancy group before extractions
(P 5 0.270, Table IV) or in the distributions of subjects
of the maxillary and mandibular correction groups (P 5
0.286 and P 5 0.365, respectively, Table V). Therefore,
the subjects of each malocclusion group were combined
in each category for other analyses.
Table VI shows that some subjects in the small over-
all ratio group moved into the normal overall ratio group
after 3 of 4 premolar extraction combinations, some
subjects in the normal overall ratio group moved into
the small or large overall ratio groups, and all subjects
in the large overall ratio group remained in the same
group after any premolar extraction combination. More-
over, Table VI shows that, in the normal overall ratio
group, Kruskal-Wallis and Bonferroni tests, and
Friedman and Tukey tests demonstrated significant dif-
ferences in the distribution of subjects with overall
tooth-size discrepancies and in the overall ratios,
respectively, between the 2 extraction combinations
including maxillary second premolars and those includ-
ing maxillary first premolars. This table also shows that,
in the large overall ratio group, there was a significant
difference in the distribution of overall ratios between
extraction combinations 3 and 4.
American Journal of Orthodontics and Dentofacial Orthopedics Endo et al 511
Volume 137, Number 4

Table IV. Distribution of subjects of overall ratio groups before extraction

Overall ratio group

Normal group Combined small and large groups Kruskal-Wallis

Group n % n % P value

Class I malocclusion 59 89.39 7 (4, 3) 10.61 (6.06, 4.55)

Class II malocclusion 63 95.45 3 (2, 1) 4.55 (3.03, 1.52) 0.270
Class III malocclusion 63 95.45 3 (2, 1) 4.55 (3.03, 1.52)
Total malocclusion 185 93.43 13 (8, 5) 6.57 (4.04, 2.53) —

Small and large overall ratio groups, respectively, in parentheses.

Table V. Distribution of subjects of maxillary and mandibular correction groups before extraction
Small Normal Large Kruskal-Wallis

Group n % n % n % P value

Maxillary correction
Class I malocclusion 20 30.30 32 48.48 14 21.21
Class II malocclusion 12 18.18 38 57.58 16 24.24 0.286
Class III malocclusion 10 15.15 41 62.12 15 22.73
Total malocclusion 42 21.21 111 56.06 45 22.73 —
Mandibular correction
Class I malocclusion 14 21.21 32 48.48 20 30.30
Class II malocclusion 14 21.21 41 62.12 11 16.67 0.365
Class III malocclusion 12 18.18 45 68.18 9 13.64
Total malocclusion 40 20.20 118 59.60 40 20.20 —

Table VII shows that some subjects in the small and
normal maxillary correction groups moved into the nor-
mal and large maxillary correction groups, respectively,
and that all or almost all subjects in the large maxillary
correction group remained in the same group with the 4
extraction combinations. Table VIII shows that all or al-
most all subjects in the small mandibular correction
group remained as they were with any extraction combi-
nation, but some subjects in the normal and large man-
dibular correction groups moved into the small and
normal groups, respectively.
Tables VII and VIII show that, in the normal max-
illary and mandibular correction groups, Kruskal-
Wallis and Bonferroni tests found significant differ-
ences in the distributions of subjects with maxillary
and mandibular corrections, respectively, between the
2 extraction combinations including maxillary second
premolars and maxillary first premolars. The tables
also show that, in the small maxillary correction and
large mandibular correction groups, significant differ-
ences in the distributions of subjects with maxillary
and mandibular corrections were found among differ-
ent extraction combinations: between combinations 1
and 4, combinations 2 and 3, and combinations 3
and 4.
Tables VII and VIII also show that in each maxillary
and mandibular correction group, Friedman and Tukey
tests demonstrated significant differences in the dis-
tributions of maxillary and mandibular corrections
between the 2 extraction combinations including maxil-
lary second premolars and maxillary first premolars.
DISCUSSION
Some evidence reflects that various tooth-size ratios
show ethnic or racial and sex differences.5,6 Our
PubMed search in March 2008 with the search subjects
‘‘tooth size ratio’’ and ‘‘tooth size discrepancy’’ found
no English references regarding the association between
Bolton overall ratio and extraction in a Japanese popu-
lation. It would, therefore, be worthwhile to examine the
effects of premolar extractions on Bolton overall ratios
and tooth-size discrepancies in Japanese people and to
compare our results with those in other populations.
Our results showing no statistically significant differ-
ences between the sexes in the overall ratio for the mal-
occlusion groups agreed with those of previous studies
on other populations.5,8,13 Some other investigations
showed statistically significant differences between
the sexes in the overall ratio for the malocclusion groups
512 Endo et al American Journal of Orthodontics and Dentofacial Orthopedics
April 2010

Table VI. Distribution of subjects of each overall ratio group after extractions and statistical comparisons
Overall ratio groups

Small Normal Large Kruskal-Wallis Bonferroni Friedman Tukey

Significant Significant
n % n % n % P value comparison P value comparison

Before extraction 8 100.00 — — — — — — — —

After extraction 4/4 5 62.50 3 37.50 0 0.00
5/5 7 87.50 1 12.50 0 0.00
0.091 0.273
4/5 4 50.00 4 50.00 0 0.00
5/4 8 100.00 0 0.00 0 0.00
Before extraction — — 185 100.00 — — — — — —
After extraction 4/4 3 1.62 90 48.65 92 49.73 4/4 vs 5/5, 4/4 vs 5/4 4/4 vs 5/5, 4/4 vs 5/4
5/5 8 4.32 125 67.57 52 28.11 5/5 vs 4/5 5/5 vs 4/5
\0.001 \0.001
4/5 1 0.54 76 41.08 108 58.38 4/5 vs 5/4 4/5 vs 5/4
5/4 16 8.65 118 63.78 51 27.57
Before extraction — — — — 5 100.00 — — — —
After extraction 4/4 0 0.00 0 0.00 5 100.00
5/5 0 0.00 0 0.00 5 100.00
— 0.017
4/5 0 0.00 0 0.00 5 100.00 4/5 vs 5/4
5/4 0 0.00 0 0.00 5 100.00

4/4, All first premolars; 5/5, all second premolars; 4/5, maxillary first and mandibular second premolars; 5/4, maxillary second and mandibular first
premolars.

Table VII. Distribution of subjects of each maxillary correction group after extractions and statistical comparisons
Maxillary correction groups

Small Normal Large Kruskal-Wallis Bonferroni Friedman Tukey

n % n % n % P value Significant comparison P value Significant comparison

Before extraction 42 100.00 — — — — — — — —

After extraction 4/4 13 30.95 29 69.05 0 0.00 4/4 vs 5/4 4/4 vs 5/5, 4/4 vs 5/4
5/5 19 45.24 23 54.76 0 0.00 5/5 vs 4/5
\0.001 \0.001
4/5 8 19.05 33 78.57 1 2.38 4/5 vs 5/4 4/5 vs 5/4
5/4 27 64.29 15 35.71 0 0.00
Before extraction — — 111 100.00 — — — — — —
After extraction 4/4 0 0.00 34 30.63 77 69.37 4/4 vs 5/5, 4/4 vs 5/4 4/4 vs 5/5, 4/4 vs 5/4
5/5 0 0.00 73 65.77 38 34.23 5/5 vs 4/5 5/5 vs 4/5
\0.001 \0.001
4/5 1 0.90 27 24.32 83 74.77 4/5 vs 5/4 4/5 vs 5/4
5/4 0 0.00 82 73.87 29 26.13
Before extraction — — — — 45 100.00 — — — —
After extraction 4/4 0 0.00 0 0.00 45 100.00 4/4 vs 5/5, 4/4 vs 5/4
5/5 0 0.00 1 2.22 44 97.78 5/5 vs 4/5
0.295 \0.001
4/5 0 0.00 0 0.00 45 100.00 4/5 vs 5/4
5/4 0 0.00 2 4.44 43 95.56

4/4, All first premolars; 5/5, all second premolars; 4/5, maxillary first and mandibular second premolars; 5/4, maxillary second and mandibular first
premolars.

among different populations.4,14,17 This permitted our
speculation that sex differences in tooth-size ratios
might be population-specific.
In this study, 2-way ANOVA showed no significant
differences in the overall ratios between the malocclu-
sion groups, but significant differences before and after
extractions (Table III). Our result that the overall ratios
decreased after extraction of any combination of premo-
lars in each malocclusion group was confirmed by Tong
et al.17 One-way ANOVA and Scheffé tests showed that
the extraction combinations with the most commonly
decreasing overall ratios in every malocclusion group
were 2 and 4; this also corresponded to the results of
the Friedman and Tukey tests: that significant differ-
ences in the distribution of overall ratios between extrac-
tion combinations 1 and 3 and combinations 2 and 4 were
found in the normal overall ratio group including most
subjects (Table VI). These findings show that the
American Journal of Orthodontics and Dentofacial Orthopedics Endo et al 513
Volume 137, Number 4

Table VIII. Distribution of subjects of each mandibular correction group after extractions and statistical comparisons
Mandibular correction groups

Small Normal Large Kruskal-Wallis Bonferroni Friedman Tukey

n % n % n % P value Significant comparison P value Significant comparison

Before extraction 40 100.00 — — — — — — — —

After extraction 4/4 40 100.00 0 0.00 0 0.00 4/4 vs 5/5, 4/4 vs 5/4
5/5 38 95.00 2 5.00 0 0.00 5/5 vs 4/5
0.136 \0.001
4/5 40 100.00 0 0.00 0 0.00 4/5 vs 5/4
5/4 37 92.50 3 7.50 0 0.00
Before extraction — — 118 100.00 — — — — — —
After extraction 4/4 70 59.32 48 40.68 0 0.00 4/4 vs 5/5, 4/4 vs 5/4 4/4 vs 5/5, 4/4 vs 5/4
5/5 28 23.73 90 76.27 0 0.00 5/5 vs 4/5 5/5 vs 4/5
\0.001 \0.001
4/5 79 66.95 39 33.05 0 0.00 4/5 vs 5/4 4/5 vs 5/4
5/4 23 19.49 95 80.51 0 0.00
Before extraction — — — — 40 100.00 — — — —
After extraction 4/4 0 0.00 30 75.00 10 25.00 4/4 vs 5/4 4/4 vs 5/5, 4/4 vs 5/4
5/5 0 0.00 22 55.00 18 45.00 5/5 vs 4/5 5/5 vs 4/5
\0.001 \0.001
4/5 1 2.50 34 85.00 5 12.50 4/5 vs 5/4 4/5 vs 5/4
5/4 0 0.00 14 35.00 26 65.00

4/4, All first premolars; 5/5, all second premolars; 4/5, maxillary first and mandibular second premolars; 5/4, maxillary second and mandibular first
premolars.

extraction of all second premolars and maxillary second
and mandibular first premolars most affect the overall ra-
tios in orthodontic treatment involving premolar extrac-
tions. However, the mean overall ratios in the Class I and
Class II malocclusion groups ranged from 87% to 89%
after these 2 extraction combinations, as shown in Table
I. Orthodontic patients with Class I and Class II maloc-
clusions treated with these 2 extraction combinations
of premolars might attain proper occlusal intercuspation.
Bolton2 stated that, after 4 premolar extractions, subjects
without a tooth-size discrepancy would have overall ra-
tios of 87% to 89%. The decrease in the overall ratio in
any extraction combinations might be because the ratios
of the mesiodistal widths of the maxillary first or second
premolars to those of the 12 maxillary teeth (15.4% or
14.3%, respectively) were smaller than those of the man-
dibular first or second premolars to the 12 mandibular
teeth (16.7% or 16.5%, respectively) in this study. In
other words, in spite of first or second premolar extrac-
tions, the rates of the sum of the mesiodistal widths of
the remaining 10 teeth to the sum of those of the 12 teeth
are smaller in the mandible than in the maxilla.
A tooth-size discrepancy can affect orthodontic
treatment outcome and its stability. The mean overall ra-
tios in every malocclusion group were more than 89%
after extraction combinations 1 and 3, demonstrating
that, if the sum of the mesiodistal widths of the remain-
ing 10 teeth in the maxilla were considered normal, that
in the mandible would be greater than the mean. From
a clinical perspective, this finding suggests that the man-
dibular incisors need to be retroclined to achieve an
excellent Class I occlusion without interproximal
enamel reduction and that the first molar relationships
need to be in Class II occlusion for an ideal occlusion
of the incisors. Moreover, it suggests that selective inter-
proximal enamel reduction might be required in the
mandibular teeth for proper occlusal interdigitation.
In this study, ratios greater than 2 SD from the Bolton
mean and discrepancies greater than 1.5 mm of the max-
illary and mandibular corrections for overall ratios were
defined as indicating clinically significant tooth-size dis-
crepancies. Our findings showed that the prevalence rate
of subjects with clinically significant tooth-size discrep-
ancies before extraction by using the Bolton standard-de-
viation definition (6.57%, Table IV) was lower than those
with the millimetric definition as shown in the distribution
of subjects with maxillary and mandibular corrections
(43.94% and 40.4%, respectively, Table V). In our study,
significant differences in the distribution of maxillary and
mandibular corrections were found in 4 extraction combi-
nations of any correction groups, whereas significant dif-
ferences in the distribution of overall ratios were found in
4 combinations and 1 combination of the normal and large
overall ratio groups, respectively. These findings sup-
ported those of some investigators who demonstrated
that the use of millimetric measurements in the correction
of tooth-size ratios could prevent clinicians from underes-
timating the prevalence of clinically significant tooth-size
discrepancies.9,13,14
Our findings showed that some subjects in the small
overall ratio group moved into the normal group, and
some subjects in the normal overall ratio group moved
514 Endo et al
into either the small or large group, but all subjects in the
large overall ratio group stayed there after premolar ex-
tractions. These findings were inconsistent with those
by Tong et al,17 who found that all the subjects in the small
overall ratio group remained in this group, but some sub-
jects in the normal and large overall ratio groups moved
into the small and normal groups, respectively. This in-
consistency might be explained by the difference in the
Bolton mean used as a criterion after extractions. We se-
lected 88%, but Tong et al17 chose 91.3%. Our findings
showed that the movements of the subjects in the normal
overall ratio group were significantly remarkable after ex-
traction combinations 1 and 3, suggesting that the tooth-
size discrepancy might be created by these 2 combina-
tions in some patients with normal overall ratios.
The maxillary and mandibular corrections for the
overall ratios could be used to evaluate a relative tooth-
size discrepancy. If tooth widths in 1 arch are considered
normal, those in the other arch will be too large or too
small, thus creating reversible, unequal, and opposite dis-
crepancies between the maxillary and mandibular
arches.14 This characteristically antipodal relationship
between maxillary and mandibular corrections was evi-
denced when we found the movements of subjects in
the small and normal maxillary correction groups corre-
sponded to those in the large and normal mandibular cor-
rection groups, respectively, and that most subjects in the
large maxillary correction group and the small mandibu-
lar correction group stayed in the same groups after ex-
tractions. In this study, the millimetric discrepancy and
the ratio discrepancy suggested that a tooth-size discrep-
ancy might be created by extraction combinations 1 and 3
in some patients with normal maxillary and mandibular
corrections; this was statistically confirmed by the Krus-
kal-Wallis and Bonferroni tests. The probable reason for
significant differences in the distribution of maxillary
and mandibular corrections between extraction combina-
tions 1 and 3, and 2 and 4, might be that the mesiodistal
width of maxillary second premolars is smaller than that
of maxillary first premolars. A part of our findings about
the changes in the distribution of subjects with millimetric
tooth-size discrepancies agreed with Saatci and Yukay,18
who found that extraction of all first premolars created
more frequent and greater discrepancies than did the other
3 extraction combinations, although they selected sub-
jects not by sex or malocclusion type.
CONCLUSIONS
In formulating a treatment plan involving premolar
extractions, orthodontists should consider that the
overall ratios might decrease, and normal and clinically
American Journal of Orthodontics and Dentofacial Orthopedics
April 2010
significant tooth-size discrepancies could change mutu-
ally after extractions.
REFERENCES
1. Bolton WA. Disharmony in tooth size and its relation to the anal-
ysis and treatment of malocclusion. Angle Orthod 1958;28:
113-30.
2. Bolton WA. The clinical application of a tooth-size analysis. Am
J Orthod 1962;48:504-29.
3. Lavelle CLB. Maxillary and mandibular tooth size in different ra-
cial groups and in different occlusal categories. Am J Orthod
1972;61:29-37.
4. Smith SS, Buschang PH, Watanabe E. Interarch tooth size rela-
tionships of 3 populations: ‘‘does Bolton’s analysis apply?’’ Am
J Orthod Dentofacial Orthop 2000;117:169-74.
5. Ta TA, Ling JKY, Hägg U. Tooth-size discrepancies among differ-
ent occlusion groups of southern Chinese children. Am J Orthod
Dentofacial Orthop 2001;120:556-8.
6. Araujo E, Souki M. Bolton anterior tooth size discrepancies
among different malocclusion groups. Angle Orthod 2003;73:
307-13.
7. Crosby DR, Alexander CG. The occurrence of tooth size discrep-
ancies among different malocclusion groups. Am J Orthod Dento-
facial Orthop 1989;95:457-61.
8. Uysal T, Sari Z, Bascifftci FA, Memili B. Intermaxillary tooth size
discrepancy and malocclusion: is there a relation? Angle Orthod
2005;75:208-13.
9. Endo T, Abe R, Kuroki H, Oka K, Shimooka S. Tooth size discrep-
ancies among different malocclusions in a Japanese orthodontic
population. Angle Orthod 2008;78:994-9.
10. Freeman JE, Maskeroni AJ, Lorton L. Frequency of Bolton tooth-
size discrepancies among orthodontic patients. Am J Orthod Den-
tofacial Orthop 1996;110:24-7.
11. Santoro M, Ayoub ME, Pardi VA, Cangialosi TJ. Mesiodistal
crown dimensions and tooth size discrepancy of the permanent
dentition of Dominican Americans. Angle Orthod 2000;70:
303-7.
12. Endo T, Shundo I, Abe R, Ishida K, Yoshino S, Shimooka S. Ap-
plicability of Bolton’s tooth size ratios to a Japanese orthodontic
population. Odontology 2007;95:57-60.
13. Othman S, Harradine N. Tooth size discrepancies in an orthodon-
tic population. Angle Orthod 2007;77:668-74.
14. Bernabe E, Major PW, Flores-Mir C. Tooth-width ratio discrep-
ancies in a sample of Peruvian adolescents. Am J Orthod Dento-
facial Orthop 2004;125:361-5.
15. Proffit WR, Ackerman JL. Contemporary orthodontics. St Louis:
C.V. Mosby; 1986. p. 123-67.
16. Vaden JL, Kiser HE. Straight talk about extraction and nonextrac-
tion: a differential diagnostic decision. Am J Orthod Dentofacial
Orthop 1996;109:445-52.
17. Tong H, Chen D, Xu L, Liu P. The effect of premolar extractions
on tooth size discrepancies. Angle Orthod 2004;74:508-11.
18. Saatci P, Yukay F. The effect of premolar extractions on tooth-
size discrepancy. Am J Orthod Dentofacial Orthop 1997;111:
428-34.
19. Gaidyte A, Baubiniene D. Influence of premolar extractions on
tooth size discrepancy. Part two: analysis of Bolton values. Stoma-
tologija 2006;8:25-9.
20. Japanese Society of Pediatric Dentistry. A study on the cephalo-
metric standards of Japanese children. Jpn J Pediatr Dent 1995;
33:659-96.