ORIGINAL ARTICLE

Effect of nonextraction and extraction

orthodontic treatments on smile esthetics
for different malocclusions
Hsin-Chung Cheng and Yi-Chun Wang
Taipei, Taiwan

Introduction: Smile esthetics is a critical factor for evaluating orthodontic treatment outcomes. In this study, we eval-
uated the differences in esthetic perceptions and smile variables between extraction and nonextraction treatments for
different malocclusions. Methods: Ninety participants were divided into 3 groups according to their pretreatment over-
jet (group I, 0-4 mm; group II, .4 mm; group III,\0 mm), with 15 extraction participants and 15 nonextraction partic-
ipants in each group. Posttreatment frontal smiling photographs were evaluated by 30 raters (10 orthodontists, 10
general dentists, 10 laypeople), and 9 smile variables were measured. Results: Smile perception for the group II
extraction subjects was higher than for the nonextraction subjects by the orthodontists and general dentists. Regard-
less of the type of treatment, group III subjects were rated lower than those in groups I and II. The arch form index,
maxillary incisor show, and smile arc were greater in the extraction participants. In multiple regression analysis, non-
extraction and group III correlated negatively with the esthetic score. Maxillary incisor show, tooth number display, and
buccal corridor ratio correlated positively with the esthetic score. Conclusions: Group II extraction subjects were
rated higher than the nonextraction subjects by dental professionals. A smile with greater maxillary incisor show,
number of displayed teeth, and buccal corridor ratio was considered more esthetic. (Am J Orthod Dentofacial
Orthop 2018;153:81-6)

S
mile esthetics has always been the focus of ortho-
dontic treatment. It has become a main reason
that patients seek orthodontic treatment,1 and
patients now evaluate their treatment outcomes not
only by the occlusion and alignment but also by the
smile esthetics.
In orthodontics, tooth extractions are a common
treatment modality. Previous studies mainly focused
on lateral profile changes regarding tooth extractions
and concluded that extraction treatment could result
in an improved lateral profile for many patients with
some combination of crowding and protrusion.2,3
For the frontal smile esthetics, authors used patients’
posttreatment frontal smiling photos for evaluation and
found no significant differences in the esthetic scores be-
tween extraction and nonextraction groups.4-9 Through a
From the School of Dentistry, College of Oral Medicine, Taipei Medical Univer-
sity; Orthodontic Division, Department of Dentistry, Taipei Medical University
Hospital, Taipei, Taiwan.
All authors have completed and submitted the ICMJE Form for Disclosure of Po-
tential Conflicts of Interest, and none were reported.
Address correspondence to: Hsin-Chung Cheng, School of Dentistry, College of
Oral Medicine, Taipei Medical University, No. 250, Wuxing Street, Xinyi District,
Taipei City 110, Taiwan; e-mail, g4808@tmu.edu.tw.
Submitted, December 2016; revised and accepted, May 2017.
0889-5406/$36.00
Ó 2017 by the American Association of Orthodontists. All rights reserved.
http://dx.doi.org/10.1016/j.ajodo.2017.05.033
comprehensive search for eligible studies, a systematic
application of eligibility criteria, and a rigorous analytical
approach, we statistically combined the data from
relevant studies. Results of our meta-analysis showed no
difference between extraction and nonextraction treat-
ments.10 A systemic review also concluded that 4 premolar
extractions and nonextraction treatment have no predict-
able effect on the overall esthetic assessment of the smile
because individual variability could influence the smile
perception as esthetically pleasing or not.11
Because previous studies regarding frontal smile es-
thetics pooled all participants for analysis without
considering the large variations among patients, they
could not determine whether extraction treatments
were more satisfactory than nonextraction treatments
(or vice versa) in terms of smile esthetics.
The purpose of this study was to compare extraction
and nonextraction treatments on smile esthetics for
different malocclusions including subjective esthetic
perceptions by panel raters and objectively by measuring
the smile variables.
MATERIAL AND METHODS
A total of 90 participants were randomly selected from
a sample of 600 patients organized in an Excel spread-
sheet (Microsoft, Redmond, Wash) from the orthodontic
81
82
department of our hospital in Taipei, Taiwan. This study
was approved by the institutional review board of Taipei
Medical University Hospital (No. 201503035). The inclu-
sion criteria for the participants were (1) all permanent
dentition, (2) completed orthodontic treatment with fixed
appliances from 2011 to 2014, and (3) a complete set of
posttreatment records, including study models, pano-
ramic radiographs, and intraoral and extraoral photo-
graphs. Participants with a large skeletal discrepancy for
which surgical orthodontic treatment might be indicated
were excluded from the study. To investigate the differ-
ence between extraction and nonextraction for different
malocclusions, the participants were further divided into
3 groups according to their pretreatment overjet: group
I (0-4 mm), group II (.4 mm), and group III (\0 mm).
Each group contained 30 participants who received
extraction (n 5 15) or nonextraction (n 5 15) treatment.
The age and sex distributions were the same in both the
extraction and nonextraction subjects in all groups.
Although the total treatment duration was longer for
the extraction than the nonextraction subjects, this differ-
ence was significant only in groups I and III.
Each participant's posttreatment frontal smiling
photograph was taken using a digital camera (550D;
Canon,) and stored in JPEG format. According to the
standard operation procedure of our department, a
well-trained photographic assistant instructed the par-
ticipants to say “7” or “cheese” while holding their heads
in a natural position. Photoshop software (Adobe Sys-
tems, San Jose, Calif) was used to manage the photo-
graphs, which were cropped to show only the perioral
area and converted to black and white images to mini-
mize the influence of other facial characteristics and
skin color. PowerPoint (Microsoft) was used to show
the photographs to the raters in a random order.
Raters, comprising 10 laypeople, 10 general dentists,
and 10 orthodontists, performed subjective evaluations
of the smile esthetics. The ages of the raters were be-
tween 30 and 50 years, and the general dentists and or-
thodontists had more than 5 years of clinical experience.
Laypeople were randomly contacted in the mass rapid
transit station. The raters had the same age and sex dis-
tributions, and no difference was noted in the years of
clinical experience between the general dentists and or-
thodontists.
Each rater used a visual analog scale to score the
smile esthetics of each photo. The scale was created on
a 100-mm uninterrupted line anchored at 0 on the left
(very unattractive) and 10 on the right (very attractive).
The raters made their decisions independently, with no
information regarding the participants. They were al-
lowed to review the slides and revise their scores until
they reached a final decision.
January 2018
Vol 153
Issue 1 American Journal of Orthodontics and Dentofacial Orthopedics
Cheng and Wang
Nine smile variables were measured from the post-
treatment frontal smiling photographs by using the
linear measurement tool in Photoshop (Fig). The tool
rounded the measurements to the nearest 0.01 mm.
Because of the differences in the magnification of the
photographs, exact linear measurements could not be
obtained. Therefore, to minimize bias, smile variables
except tooth number display and midline were measured
as ratios.
Seven smile variables were ratios (Fig): (1) smile arc
ratio, distance of the maxillary incisor edge to the inter-
canine connecting line divided by the distance of the
lower lip to the intercanine connecting line; (2) maxillary
incisor show, distance of the maxillary incisal edge to the
upper lip divided by the incisor width; (3) mandibular
teeth exposure, visible mandibular incisor length divided
by the mandibular incisor width; (4) arch form index, in-
tercanine width divided by intermolar width; (5) buccal
corridor ratio, intercommissure width divided by interca-
nine width; (6) smile index, intercommissure width
divided by the interlabial gap; and (7) interlabial gap, in-
terlabial gap divided by intercanine width.
Two smile variables were not ratios: (1) midline, up-
per and lower dental midlines (on, 1; off, 0) and (2) tooth
number display: exposed maxillary teeth.
Statistical analysis
Statistical analyses were performed using the R Data
Analysis and Guiding System (Chinese Association of R
Software Research and Application, Taiwan). We
randomly assigned the 30 participants in each group
into 2 treatments (each treatment had 150 measure-
ments), which provided 87% power to detect a differ-
ence between means at a significance level of 5% by
using a 2-sided t test. A power test was performed to
ensure an adequate sample size. A 2-sample t test was
used to compare the esthetic scores and smile variables
between the extraction and nonextraction subjects in
each group. One-way analysis of variance was used to
compare smile perceptions among the 3 types of raters.
Multiple regression analysis was used to evaluate the ef-
fects of tooth extraction treatment and groups on the
smile esthetic score and to identify whether any variables
influenced the smile esthetic scores. The level of signif-
icance was established as P \0.05 for all statistical tests.
RESULTS
Table I shows the mean esthetic scores of the extrac-
tion and nonextraction subjects stratified by group. In
group II, extraction was rated higher than nonextraction.
No differences were observed in groups I and III. More-
over, regardless of the type of treatment, the group III
Cheng and Wang 83

Fig. Smile variables: A, smile arc ratio (a, distance of maxillary incisor edge to intercanine connecting
line; b, distance of lower lip to the intercanine connecting line); B, maxillary incisor show (c, distance of
maxillary incisal edge to upper lip; d, maxillary incisor width); C, mandibular teeth exposure (e, visible
mandibular incisor length; f, mandibular incisor width); D, arch form index (g, intercanine width; h, inter-
molar width); E, buccal corridor ratio (i, intercommissure width; j, intercanine width); F, smile index (i,
intercommissure width; k, interlabial gap); G, interlabial gap (k, interlabial gap; j, intercanine width).

participants received significantly lower ratings than did

Table I. Esthetic scores of extraction and nonextrac-
the groups I and II participants.
tion stratified by group
The esthetic perceptions of the different raters were
compared (Table II). Group II extraction subjects were Extraction Nonextraction
rated higher than the nonextraction subjects by ortho- Mean 6 SD Mean 6 SD P value
dontists and general dentists. Therefore, the total Group I 53.19 6 16.58 52.14 6 17.25 0.35
esthetic score of extraction was higher. In all groups, Group II 53.69 6 16.73 48.81 6 17.90 \0.001*
the ratings by laypeople showed no difference between Group III 48.90 6 16.30 47.76 6 16.67 0.331
extraction and nonextraction.
*P \0.001.
Table III shows the smile variables of extraction and
nonextraction stratified by group. In group I, smile arc,
maxillary incisor show, and arch form index were greater
in extraction subjects. In group III, arch form index was the laypeople. For the participants who had extraction
greater in the extraction group. In group II, no difference treatment, tooth number display correlated positively
was observed between extraction and nonextraction. A with esthetic scores by all raters, and mandibular teeth
comparison of the smile variables of the 3 groups exposure correlated negatively with esthetic scores by
showed that maxillary incisor show was significantly the laypeople. For the participants who had nonextrac-
less in group III than in group II, and lower lip exposure tion treatment, buccal corridor ratio correlated positively
was significantly greater in group III than in group II. with esthetic scores by the general dentists and
Multiple regression analysis was used to evaluate the laypeople.
effect of treatments and groups on the smile esthetic
scores (Table IV). Nonextraction and group III correlated DISCUSSION
negatively with the smile esthetic score. This study focused on frontal smile esthetics. One
Table V shows the results for the multiple regression important factor that influences frontal smile esthetics
analysis of the esthetic scores and smile variables strati- is different overjets of anterior teeth. The participants
fied by rater type. For all participants, maxillary incisor in this study were divided into 3 groups according to
show correlated positively with esthetic scores by the or- their pretreatment overjet. Instead of the ANB angle,
thodontists; tooth number display and maxillary incisor overjet was used for patient classification because (1)
show correlated positively with esthetic scores by the the ANB angle of cephalometric measurements may be
general dentists; and buccal corridor ratio and maxillary influenced by the diversity of individual cranial bases,
incisor show correlated positively with esthetic scores by and (2) the skeletal measurements of cephalograms do

American Journal of Orthodontics and Dentofacial Orthopedics January 2018
Vol 153
Issue 1
84 Cheng and Wang

Table II. Esthetic perceptions of the raters stratified by group

Orthodontists General dentists Laypeople

Extraction Nonextraction Extraction Nonextraction Extraction Nonextraction

Mean 6 SD Mean 6 SD P value Mean 6 SD Mean 6 SD P value Mean 6 SD Mean 6 SD P value

Group I 56.07 6 15.52 53.85 6 14.58 0.20 52.92 6 13.87 52.08 6 14.62 0.61 50.59 6 19.49 50.48 6 21.56 0.96
Group II 56.69 6 14.42 49.73 6 14.14 \0.001y 52.29 6 13.56 47.71 6 13.96 0.004* 52.1 6 20.93 49 6 20.10 0.19
Group III 50.47 6 16.59 50.25 6 13.87 0.90 49.65 6 16.11 46.60 6 14.50 0.90 46.56 6 20.55 46.41 6 20.61 0.95

*P \0.01; yP \0.001.

Table III. Smile variables of extraction and nonextraction stratified by group

Group I Group II Group III

Extraction No-ext Extraction Nonextraction Extraction Nonextraction

Mean 6 SD Mean 6 SD P value Mean 6 SD Mean 6 SD P value Mean 6 SD Mean 6 SD P value

Smile arc 0.52 6 0.27 0.29 6 0.25 0.01* 0.53 6 0.33 0.5 6 0.34 0.75 0.42 6 0.28 0.35 6 0.36 0.14
Tooth number 8.53 6 1.60 8.53 6 0.92 0.78 8.67 6 1.45 8.67 6 1.80 0.91 7.87 6 1.77 8.13 6 1.77 0.64
Maxillary incisor 1.01 6 0.32 0.76 6 0.29 0.04* 0.99 6 0.26 0.93 6 0.31 0.53 0.81 6 0.25 0.69 6 0.24 0.15
Midline 0.60 6 0.51 0.73 6 0.46 0.45 0.80 6 0.41 0.67 6 0.49 0.42 0.73 6 0.46 0.73 6 0.46 1.00
Buccal corridor 1.54 6 0.14 1.62 6 0.15 0.13 1.68 6 0.18 1.68 6 0.12 0.79 1.61 6 0.12 1.58 6 0.14 0.52
Smile index 5.70 6 1.09 5.97 6 1.91 0.98 6.37 6 1.26 6.51 6 2.44 0.68 6.54 6 3.25 6.91 6 2.15 0.30
Arch form 0.83 6 0.04 0.76 6 0.02 \0.001y 0.81 6 0.05 0.78 6 0.04 0.06 0.82 6 0.05 0.75 6 0.05 \0.001y
Mandibular teeth 0.41 6 0.24 0.69 6 0.40 0.06 0.37 6 0.35 0.42 6 0.39 0.80 0.71 6 0.42 0.57 6 0.41 0.37
Interlabial gap 0.28 6 0.07 0.29 6 0.08 0.43 0.27 6 0.06 0.29 6 0.09 0.71 0.28 6 0.09 0.25 6 0.09 0.29

*P \0.05; yP \0.001.

Peck et al12 indicated that dental professionals and

Table IV. Multiple regression analysis of the esthetic
scores for nonextraction treatment and group III
Estimate SD P value
Nonextraction 2.358 0.652 \0.001*
Group III 4.339 0.798 \0.001*
*P \0.001.
not always reflect the real condition of a dental occlu-
sion.
Previous studies comparing smile esthetics between
extraction and nonextraction reported no differences
between the 2 treatments.4-9 Because the participants
in the study were pooled for the analysis, the effect of
the treatment for different malocclusions could not be
investigated. We divided the participants into 3
groups. Results show that for group II, the
orthodontists and general dentists rated extraction
significantly higher than nonextraction. Regardless of
the treatment types, group III had significantly lower
esthetic scores than did groups I and II. This result may
be due to less maxillary incisor show and greater
mandibular teeth exposure resulting from the dental
compensation for group III.
laypeople judge facial esthetics differently because den-
tists are trained to observe features that are not obvious
to the general public. In this study, the smile ratings
given by laypeople did not differ significantly between
the extraction and nonextraction subjects in any group.
This result is consistent with that of Ghaffar and Fida.5
The arch form index, measured as the ratio of inter-
canine width to intermolar width, was significantly
higher in the extraction treatment of groups I and III.
Closure of the extraction space often results in mesial
movement of the molars into a narrower arch and distal
movement of the canines into a wider arch, causing the
ratio to be greater. Maximum anchorage is often
required for group II extraction treatment with little or
no molar mesial movement; this might be why no differ-
ence was observed in the arch form index between
extraction and nonextraction in group II. Moreover,
although the arch form index was higher in extraction
than nonextraction in groups I and III, no difference
was observed in the buccal corridor ratio. This result is
consistent with a previous study that reported that
although significant arch width changes were likely to
occur, no clinically relevant effects on the buccal
corridor ratio were identified.13

January 2018
Vol 153
Issue 1 American Journal of Orthodontics and Dentofacial Orthopedics
Cheng and Wang 85

no difference in the tooth number display between extrac-

Table V. Multiple regression analysis of smile vari-
tion and nonextraction groups. However, smiles that
ables and esthetic scores stratified by rater
display more teeth are considered to be more esthetic.
Estimate SD P value In the nonextraction participants, the buccal corridor
All raters ratio correlated positively with the esthetic scores by all
Total raters except for orthodontists, indicating that a full smile
Maxillary incisor 6.51 2.30 0.006y
is unnatural and less attractive to general dentists and
Buccal corridor 10.48 4.56 0.02*
Extraction laypeople. Orthodontists reported no esthetic preference
Tooth number 2.4 0.60 \0.001z in the buccal corridor ratios between the 2 treatments;
Nonextraction this is consistent with the results of previous studies.16,17
Buccal corridor 15.62 6.38 0.02* There were some limitations in this study. First,
Orthodontists
although the participants were divided into 3 groups ac-
Total
Maxillary incisor 7.43 2.31 0.004y cording to their pretreatment overjet, there were still in-
Extraction dividual variations such as differences in space
Tooth number 2.48 0.67 \0.001z deficiency among each group. Second, smile esthetics
Nonextraction is dynamic and difficult to measure; there is no standard
General dentists
method for evaluation. Third, although the cropped
Total
Maxillary incisor 4.77 2.17 0.03* photographs reduced the possibility of grading smiles
Tooth number 1.14 0.44 0.01* according to criteria that were not under orthodontic
Extraction control, smiles are judged according to the balance of
Tooth number 2.27 0.55 \0.001z the whole face in actual life.18 For further research, it
Nonextraction
is better to achieve equal baseline characteristics of par-
Buccal corridor 12.69 5.56 0.03*
Laypeople ticipants, and digital videos for 3-dimensional views of
Total the smiles may be helpful to obtain reproducible and
Maxillary incisor 6.67 2.94 0.03* comprehensive smiling photographs.
Buccal corridor 19.71 5.83 0.001z The results of this study indicate that smiles with
Extraction
greater maxillary incisor show, tooth number display,
Tooth number 2.74 0.73 \0.001z
Mandibular teeth 7.81 2.93 0.01* and buccal corridor ratio are more esthetic. In addition,
Nonextraction extraction treatments tend to retrude the lips and retract
Buccal corridor 23.32 9.11 0.01* the incisors, increasing the maxillary incisor show,
*P \0.05; yP \0.01; zP \0.001.
whereas nonextraction treatments tend to flare the inci-
sors, flatten the smile arc, and protrude the lips.

The maxillary incisor show and smile arc were signif-
icantly smaller in group I nonextraction subjects. Nonex-
traction orthodontic treatment by means of expansion
of the dental arch with increased maxillary incisor torque
might flatten the smile arc and reduce the incisor display.
Nonextraction treatment and group III correlated
negatively with the smile esthetic scores, implying that
nonextraction treatment and group III participants
received lower esthetic scores. For all participants, the
maxillary incisor show correlated positively with the
esthetic scores by all raters, indicating that a greater
maxillary incisor show results in a more esthetic smile.
This finding is similar to previous studies that have
shown that a smile with a full incisor display is deemed
more youthful and esthetic.14,15
Although the participants in extraction and nonextrac-
tion displayed an equal number of teeth in all groups when
smiling, the numbers of tooth display correlated positively
with the esthetic scores for extraction treatment. This
result is similar to that of Kim and Gianelly,7 who reported
CONCLUSIONS
1. The esthetic scores by dental professionals were
higher in extraction than nonextraction in group
II. Laypeople had no esthetic preference regarding
the type of treatment.
2. The smile esthetic score of group III was signifi-
cantly lower than the scores for groups I and II
because of the smaller maxillary incisor show and
greater mandibular teeth exposure.
3. Regarding the smile variables, the arch form index
was higher in the groups I and III extraction partic-
ipants. The smile arc and maxillary incisor show
were greater in the group I extraction subjects.
4. The maxillary incisor show correlated positively with
the esthetic score in all participants. The tooth num-
ber display correlated positively with the esthetic
score in the extraction participants. The buccal
corridor ratio correlated positively with the esthetic
score in the nonextraction participants.

American Journal of Orthodontics and Dentofacial Orthopedics January 2018
Vol 153
Issue 1
86
REFERENCES
1. Berg R. Orthodontic treatment—yes or no? A difficult decision in
some cases. A contribution to the discussion. J Orofac Orthop
2001;62:410-21.
2. Bowman SJ, Johnston LE Jr. The esthetic impact of extraction and
nonextraction treatments on Caucasian patients. Angle Orthod
2000;70:3-10.
3. Paquette DE, Beattie JR, Johnston LE Jr. A long-term comparison
of nonextraction and premolar extraction edgewise therapy in
“borderline” Class II patients. Am J Orthod Dentofacial Orthop
1991;102:1-14.
4. Johnson DK, Smith RJ. Smile esthetics after orthodontic treatment
with and without extraction of four first premolars. Am J Orthod
Dentofacial Orthop 1995;108:162-7.
5. Ghaffar F, Fida M. Effect of extraction of first four premolars on
smile aesthetics. Eur J Orthod 2011;33:679-83.
6. Meyer AH, Woods MG, Manton DJ. Maxillary arch width and
buccal corridor changes with orthodontic treatment. Part 2: attrac-
tiveness of the frontal facial smile in extraction and nonextraction
outcomes. Am J Orthod Dentofacial Orthop 2014;145:296-304.
7. Kim E, Gianelly A. Extraction vs nonextraction: arch widths and
smile esthetics. Angle Orthod 2003;73:354-8.
8. Isiksal E, Hazar S, Akyalcin S. Smile esthetics: perception and com-
parison of treated and untreated smiles. Am J Orthod Dentofacial
Orthop 2006;129:8-16.
9. Shafiee R, Korn EL, Pearson H, Boyd RL, Baumrind S. Evaluation of
facial attractiveness from end-of-treatment facial photographs.
Am J Orthod Dentofacial Orthop 2008;133:500-8.
January 2018
Vol 153
Issue 1 American Journal of Orthodontics and Dentofacial Orthopedics
Cheng and Wang
10. Cheng HC, Wang YC, Tam KW, Yen MF. Effects of tooth extraction
on smile esthetics and the buccal corridor: a meta-analysis. J Dent
Sci 2016;11:387-93.
11. Janson G, Branco NC, Fernandes TM, Sathler R, Garib D, Lauris JR.
Influence of orthodontic treatment, midline position, buccal
corridor and smile arc on smile attractiveness. Angle Orthod
2011;81:155-63.
12. Peck S, Peck L, Kataja M. Some vertical lineaments of lip position.
Am J Orthod Dentofacial Orthop 1992;101:519-24.
13. Meyer AH, Woods MG, Manton DJ. Maxillary arch width and
buccal corridor changes with orthodontic treatment. Part 1:
differences between premolar extraction and nonextraction
treatment outcomes. Am J Orthod Dentofacial Orthop 2014;
145:207-16.
14. Dong JK, Jin TH, Cho HW, Oh SC. The esthetics of the smile: a re-
view of some recent studies. Int J Prosthodont 1999;12:9-19.
15. Hulsey CM. An esthetic evaluation of lip-teeth relationships pre-
sent in the smile. Am J Orthod 1970;57:132-44.
16. McNamara L, McNamara J Jr, Ackerman M, Baccetti T. Hard- and
soft-tissue contributions to the esthetics of the posed smile in
growing patients seeking orthodontic treatment. Am J Orthod
Dentofacial Orthop 2008;133:491-9.
17. Roden-Johnson D, Gallerano R, English J. The effects of buccal
corridor spaces and arch form on smile esthetics. Am J Orthod Den-
tofacial Orthop 2005;127:343-50.
18. Rigsbee OH 3rd, Sperry TP, BeGole EA. The influence of facial an-
imation on smile characteristics. Int J Adult Orthod Orthognath
Surg 1988;3:233-9.