ORIGINAL ARTICLE

Extraction of maxillary first permanent molars

in patients with Class II Division 1 malocclusion
Mattijs J. P. Stalpers,a Johan W. Booij,b Ewald M. Bronkhorst,c Anne Marie Kuijpers-Jagtman,d
and Christos Katsarose
Nijmegen and Gorinchem, the Netherlands

Introduction: Our objectives were to assess treatment outcomes in Class II Division 1 patients who were
treated orthodontically with extraction of the maxillary first permanent molars and to describe the changes
in their facial profiles. Methods: This was a prospective, longitudinal, 1-group outcome analysis in a private
practice, with outcome evaluation by independent observers at an academic clinic. One hundred consec-
utively treated patients were enrolled prospectively and treated by 1 orthodontist. The inclusion criteria were
white, Class II Division 1, sagittal overjet of ⱖ4 mm, extraction of maxillary first permanent molars, no missing
teeth or agenesis, maxillary third molars present, and 1-stage full fixed appliance treatment. Standardized
lateral cephalometric radiographs were made before and after active treatment. Occlusal outcome was
scored on dental casts by comparing pretreatment and posttreatment casts with the peer assessment rating
(PAR) index. Backward regression analysis was used to explain the soft-tissue changes on the basis of dental
changes and the soft-tissue characteristics. Results: The mean reduction in weighted PAR score was 89.9%
(SD, 0.9). During treatment, the lower lip retruded 1.6 mm (SD, 1.7) relative to the esthetic line. The nasolabial
angle became 2.1° (SD, 7.0) more obtuse during treatment. Overjet reduction and initial upper lip thickness
could explain 15% of the variation in upper lip position. The changes in the position of the mandibular incisor
relative to the Point A-pogonion line and initial lower lip thickness could explain 23% of the variation of lower
lip position. Conclusions: Orthodontic treatment involving extraction of the maxillary first permanent molars
has a good treatment outcome. Extraction of the maxillary first permanent molars has only a small effect on
the soft-tissue profile. (Am J Orthod Dentofacial Orthop 2007;132:316-23)

T
he Class II relationship is the most prevalent
subclassification of malocclusion. Recognizing
the limitations of this description, epidemiolo-
gists have formulated indexes and scales for use in
public health studies. The results of these surveys
suggest that 15% to 22% of American teenagers have
an overjet of 5 mm or greater.1 For those seeking
orthodontic treatment to have the overjet corrected,
there are generally 3 treatment options. Clinical and
anamnestic findings are key factors in narrowing the
choices. Growth modification, with either headgear or
activator, might be suggested for a prepubertal patient
a
Orthodontist, Department of Orthodontics and Oral Biology, Radboud Uni-
versity Nijmegen Medical Center, Nijmegen, the Netherlands.
b
Private practice, Gorinchem, the Netherlands.
c
Biostatistician, Department of Preventive and Curative Dentistry, Radboud
University Nijmegen Medical Center, Nijmegen, the Netherlands.
d
Professor and chair, Department of Orthodontics and Oral Biology, Radboud
University Nijmegen Medical Center, Nijmegen, the Netherlands.
e
Professor, Department of Orthodontics and Oral Biology, Radboud University
Nijmegen Medical Center, Nijmegen, the Netherlands.
Reprint requests to: Christos Katsaros, Radboud University Nijmegen
Medical Center, Department of Orthodontics and Oral Biology, 117
Tandheelkunde, PO Box 9101, 6500 HB Nijmegen, the Netherlands; e-mail,
c.katsaros@dent.umcn.nl.
Submitted, June 2005; revised and accepted, January 2006.
0889-5406/$32.00
Copyright © 2007 by the American Association of Orthodontists.
doi:10.1016/j.ajodo.2006.01.034
316
with a favorable growth pattern. For a postpubertal
patient seeking treatment, surgical correction of the jaw
abnormality can be an option. The third option is
orthodontic treatment combined with extraction of
teeth. In case of crowding in both arches, extraction of
4 teeth is often indicated. In case of overjet and a good
or potentially good mandibular arch, extractions can be
limited to the maxillary arch only. Space analysis of the
mandibular arch involves factors such as crowding or
spacing, the curve of Spee, and the position of the
mandibular incisors.2 Within reason, these incisors can
be left slightly proclined. This compensated position is
presumably stable, provides extra arch space, and
reduces the amount of retraction needed for the maxil-
lary incisors. This in turn can provide an improved
facial profile in many Class II patients.3
When the clinician decides to extract teeth, the next
question is which teeth are to be extracted. The most
obvious choice might be to extract the first or second
premolars. In certain cases, extraction of the first perma-
nent molars can be preferred. In addition to extracting the
first permanent molars in a systematic othodontic treat-
ment approach, there are certain objective indications for
first molar extractions. These might be extensive caries
lesions, large fillings, endodontic or periodontal problems,
American Journal of Orthodontics and Dentofacial Orthopedics
Volume 132, Number 3
or hypoplastic enamel. A dogma in orthodontics is that
extraction in the posterior part of the dental arch has a
bite-closing effect that would be beneficial for high-angle
patients and, on the other hand, detrimental in low-angle
patients.4 Another extensively debated belief is that ex-
traction in the molar area would have less influence on the
facial profile than premolar extractions. In a study of
cephalometric changes in a group of borderline extraction
patients treated with or without premolar extractions, it
was observed that the facial profiles of the extraction
group were, on average, 2 mm flatter than the profiles of
the nonextraction patients.5 Katsaros et al6 found, in a
study on extraction vs nonextraction treatment, high vari-
ability in soft-tissue changes, which were unpredictable
for an individual patient.6 In the orthodontic literature,
there are few data concerning the extraction of first molars
as part of orthodontic treatment. Patients and methods
have been described, but prospectively collected data are
absent.7,8
The purposes of our study were to assess treatment
outcomes in Class II Division 1 patients treated with
extraction of the maxillary first permanent molars and
to describe the changes in their facial profiles.
MATERIAL AND METHODS
The sample consisted of 100 consecutively treated
patients (45 girls, 55 boys) who were enrolled prospec-
tively. The intake of patients started in 1998 and
finished in 2004. All patients were treated by 1 orth-
odontist (J.W.B.). The following inclusion criteria were
used: white, Class II Division 1 malocclusion, sagittal
overjet of ⱖ4 mm, treatment plan included extraction
of maxillary first permanent molars, no missing teeth or
agenesis, maxillary third molars present, and 1-stage
full fixed appliance treatment. The mean age at the start
of treatment was 13.2 years (range, 10.5-17.2; SD, 1.4).
The maxillary second molars and the mandibular
molars were banded before the extractions. The molar
bands had 6-mm single 0.018-in round buccal tubes and
palatal sheaths. The maxillary first molars were sepa-
rated to facilitate the extractions, and, after 3 weeks,
both arches were bonded with light wire Begg brackets.
The maxillary premolars were not bonded in the first
phase of the treatment to reduce friction. The maxillary
second molars were connected by a removable palatal
bar to increase anchorage and to correct possible
rotations and transverse malpositions. Canine retraction
was performed with elastics (light 5/16 in) from the
maxillary canine to the maxillary second molar, and the
patient was instructed to replace them only once a
week. If the maxillary second molars moved mesially
too fast, the intra-arch elastics were replaced by Class II
elastics. In most patients, Class I canine and premolar
Stalpers et al 317
interdigitation was reached in 6 months. At this point,
the maxillary premolars were bonded. In the end phase
of treatment, adjustments were made in the archwires
for detailed finishing. In most cases, retention was with
fixed retainers. To prevent overeruption of the mandib-
ular second molars, local retention wires were bonded
between the first and second molars; the wires were
removed when the maxillary third molars were in
occlusal contact with the mandibular second molars.
Standardized lateral cephalometric radiographs
were made before (T1) and after (T2) active treatment.
The radiographs were scanned on an 8-bit scanner
(Linotype-Hell, Eschborn, Germany). Tracing of the
films was done in Viewbox (dHal Orthodontic Soft-
ware, Athens, Greece), a software program for cepha-
lometric analysis. All tracings were corrected to life
size. Skeletal, dental, and soft-tissue cephalometric
landmarks were used. The cephalometric landmarks
and reference lines are illustrated in Figure 1.
To evaluate the influence of treatment on different
vertical facial types, the sample was divided into 3
groups: horizontal (ANS-Me/N-Me ⱕ56%; n ⫽ 18),
normal (56% ⬍ANS-Me/N-Me ⬍58%; n ⫽ 21), and
divergent (ANS-Me/N-Me ⱖ58%; n ⫽ 61).9
Mean tracings of the 100 patients were constructed
and superimposed. To calculate a mean tracing, the
tracings were superimposed on the anterior nasal spine
(ANS)-posterior nasal spine (PNS) line at ANS. After
all tracings had been superimposed on each other,
Viewbox calculated the mean coordinates for each
cephalometric point. These mean coordinates were
used to construct a new tracing, the mean tracing. This
was done for T1 and T2. These mean tracings were, in
their turn, superimposed on ANS-PNS at ANS to
illustrate treatment effect and growth.
The occlusal outcome of treatment was scored on
the dental casts, by comparing the casts at T1 and T2
with the peer assessment rating (PAR index).10 The
PAR index is applied to a patient’s pretreatment and
posttreatment dental casts. Scores are assigned to the
various occlusal traits that make up the malocclusion.
The components of the PAR index are alignment of
maxillary anterior segments, alignment of mandibular
anterior segments, left buccal occlusion, right buccal
occlusion, overjet, overbite, and center line. British
weightings were used; this means that the individual
scores for overjet were multiplied by 6, overbite by 2,
and center line by 4. A score of zero indicates good
alignment and occlusion, and higher scores indicate
increased levels of irregularity or malocclusion. The
difference between the pretreatment and posttreatment
PAR scores indicates improvement as a result of
orthodontic intervention. Furthermore, the percentage
318 Stalpers et al
Fig 1. Cephalometric points and reference lines used in
study.
of improvement was calculated. This percentage also
reflects the change relative to the pretreatment score but
gives a more sensitive assessment.
To illustrate the degree of improvement, the nomo-
gram was used, in which the degree of change is
separated into 3 sections: worse or no different (patients
with less than 30% reduction), improved (patients with
American Journal of Orthodontics and Dentofacial Orthopedics
September 2007
ⱖ30% reduction), and greatly improved (generally a
reduction of 22 weighted PAR points or more).11
To assess the method error, the cephalograms of 20
patients and the dental casts of 50 patients were
measured by a second orthodontist. For the assessment
of interobserver error, the duplicate measurement error
was calculated with Dahlberg’s formula.12 Paired t tests
were used to assess systematic differences between the
2 observers for the cephalometric variables.
Statistical analysis
Differences between the cephalometric variables
and the PAR score before and after treatment were
analyzed by using paired t tests. To evaluate the
influence of treatment on the various vertical facial
types, ANOVA was applied, followed post hoc by the
Bonferroni correction for multiple testing. T tests were
used to assess the difference in the treatment effect
between boys and girls. For only 2 of the 22 variables
tested (21 cephalometric variables and the PAR score),
a statistically significant difference between the sexes
was found: overbite and length of the nose (N-No).
Therefore, the results are presented for the complete
patient group, regardless of sex.
Multiple linear regression was used to attempt to
explain the soft-tissue changes on the basis of dental
changes. The change in the sagittal position of the
upper lip relative to the esthetic line (Ls to E-line) was
used as the dependent variable. Several independent
variables were used: change in the sagittal position of
the maxillary incisor relative to Point A-pogonion (U1
to A-Pog), change in the maxillary incisor inclination
(U1L/ANS-PNS), change in overjet, and sex. The
pretreatment thickness (Ls-U1) of the upper lip was
added as a controlling variable. To reduce the initial
model, stepwise backward regression was applied with
a threshold for the P value of .1 for removing a variable
from the model. A similar approach was used for the
lower lip: change in the sagittal position of the lower lip
relative to the esthetic line (Li to E-line) was used as a
dependent variable, and change in the sagittal position
of the mandibular incisor relative to A-Pog (L1 to
A-Pog), change in mandibular incisor inclination (L1L/
ML), change in overjet, and sex were the independent
variables. Again, the pretreatment thickness of the lip
(Li-L1) was added as a controlling variable.
To analyze the relationship between age at start of
treatment and treatment effect on cephalometric vari-
ables, scatter plots were made to check for nonlinear
relationships. Without indications for a nonlinear rela-
tionship, Pearson correlations were calculated to quan-
tify the relationship between age and the cephalometric
variables. For the relationshipbetween the reduction in
American Journal of Orthodontics and Dentofacial Orthopedics Stalpers et al 319
Volume 132, Number 3

Table I. Results of method error analysis (first 2 col- score was 89.9% (SD, 0.9). The mean absolute reduc-
umns relate to paired t test) tion was 26.3 points (SD, 7.3) to 2.9 points (SD, 2.4) at
Mean difference Duplicate
T2. Thirteen patients had a PAR score of zero at T2; 10
between measurement was the highest PAR score at T2. There were no
Variable observers P value error patients in the “worse or no different” group, 27 were in
the “improved” group, and 73 were in the “greatly
Dental cast analysis
PAR ⫺1.595 ⬍.0001 3.297 improved” group. The nomogram is shown in Figure 2.
Cephalometric analysis For both the absolute and the relative reduction of
SNA angle (°) ⫺0.044 .357 0.194 the PAR score, no statistically significant correlation
SNB angle (°) ⫺0.047 .147 0.133 with age was found.
ANB angle (°) 0.015 .682 0.145
Just 1 cephalometric variable showed a statistically
SN/ANS-PNS (°) 0.065 .017 0.114
SN/ML (°) ⫺0.044 .176 0.133 significant correlation with age. The difference of N-No
ANS-PNS/ML (°) ⫺0.106 .006 0.165 (T2 – T1) had a negative correlation with age (r ⫽
ANS-ME/N-ME (ratio) 0.094 .443 0.495 – 0.344, P ⫽ .0%). All other cephalometric variables
U1L/ANS-PNS (°) ⫺0.032 .583 0.238 had P values above 5% for the statistical significance of
U1 to A-Pog (mm) 0.121 .002 0.167
the correlation with age. Therefore, the results are
L1L/ML (°) 0.124 .043 0.254
L1 to A-Pog (mm) 0.121 .005 0.184 presented without considering the patients’ ages.
Overbite (mm) ⫺0.018 .338 0.068 As shown in Table II, at T1, the patient group had
Overjet (mm) 0.003 .861 0.075 typical Class II Division 1 characteristics: enlarged
Nasolabial angle (°) 0.759 .358 3.351 ANB angle (mean, 5.5°), overjet (mean, 7.3 mm), and
Ls to Sn-Pog’ (mm) ⫺0.041 .657 0.375
overbite (mean, 2.9 mm). Furthermore, the maxillary
Li to Sn-Pog’ (mm) ⫺0.076 .387 0.358
Ls to E-line (mm) ⫺0.029 .658 0.268 incisors were protruded in relation to the A-Pog line. At
Li to E-line (mm) ⫺0.059 .442 0.310 T2, the ANB angle showed a significant reduction of
N-No (mm) ⫺0.144 .185 0.444 1.8° (SD, 1.5), mainly due to a reduction in the SNA
Ls-U1 (mm) 0.047 .702 0.497 angle.
Li-L1 (mm) ⫺0.206 .024 0.383
Superimposition of the mean tracings at T1 and T2
visually represents the overall treatment effect (Fig 3).
During treatment, the maxillary incisors were re-
the PAR score (both absolute and relative reductions), tracted 2.7 mm (SD, 1.9) in relation to A-Pog. The
the same procedure was followed. inclination of the maxillary and mandibular incisors
changed significantly. The maxillary incisors retro-
RESULTS clined 2.3° (SD, 6.6) on ANS-PNS, and the mandibular
The average duration of active orthodontic treat- incisors proclined 5.3° (SD, 5.0) to 103.2°(SD, 6.7) on
ment was 2.5 years (SD, 0.6). All 100 patients who mandibular line. The retraction and the retroclination of
were enrolled in the study finished treatment. The mean the maxillary incisors and the proclination of the
age at T2 was 15.7 years (SD, 1.6; range, 12.4-19.8). mandibular incisors together with growth must have
In Table I, for all variables, the results of the paired contributed to the overjet reduction of 4.8 mm (SD, 1.8)
t test comparing the 2 observers and the duplicate from 7.3 mm (SD, 1.8) at T1 to 2.5 mm (SD, 0.9) at T2.
measurement error are given. For 6 cephalometric Overbite was reduced from 2.9 mm (SD, 2.7) at T1 to
variables (SN/ANS-PNS, ANS-PNS/ML, U1 to A-Pog, 1.3 mm (SD, 0.8) at T2. These changes were all
L1L/ML, L1 to A-Pog, and Li-L1) and the PAR score, significant.
a statistically significant difference between the 2 The position of the upper lip relative to the E-line
observers was found. For angles, a maximum duplicate (Ls to E-line) was 0.77 mm (SD, 2.5) at T1. At T2, its
measurement error of 0.25° and, for distances, a max- position was 3.3 mm (SD, 2.1) behind the E-line, a
imum duplicate measurement error of 0.38 mm were change of –2.6 mm (SD, 1.7). The position of the lower
found. However, for all variables, the duplicate mea- lip relative to the E-line (Li to E-line) was ⫹0.1 mm
surement error was small compared with the standard (SD, 2.3) at T1. At T2, its position was 1.6 mm (SD,
deviations of the measurement (Table II). 2.3) behind the E-line, a change of –1.6 mm (SD, 1.7).
The descriptive statistics for the PAR score are The nasolabial angle became 2.1° (SD, 7.0) more
summarized in Table II. The average PAR score at T1 obtuse during treatment.
was 29.2 (SD, 7.3). Eighty-four of the 100 patients had In Table III, the results of the regression analysis
an initial score of ⱖ22 points; 15 was the lowest initial aiming at prediction of the soft-tissue changes by dental
score. The mean percentage reduction in weighted PAR variables are given. For the upper lip after applying
320 Stalpers et al American Journal of Orthodontics and Dentofacial Orthopedics
September 2007

Table II. Descriptive statistics for cephalometric and dental cast measurements (P values calculated with paired t tests)
T1 T2 T2–T1

Variable Mean SD Mean SD Mean SD P value

Dental cast analysis

PAR* 29.16 7.34 2.88 2.43 ⫺26.28 7.32 ⬍.0001
PAR† 29.16 7.34 2.88 2.43 89.9% 0.9 ⬍.0001
Cephalometric analysis
Skeletal sagittal
SNA angle (°) 79.82 3.74 77.70 3.94 ⫺2.12 1.92 ⬍.0001
SNB angle (°) 74.34 3.67 73.98 3.72 ⫺0.35 1.24 .0051
ANB angle (°) 5.48 1.92 3.72 2.23 ⫺1.76 1.54 ⬍.0001
Skeletal vertical
SN/ANS-PNS (°) 7.23 3.31 7.30 3.38 0.07 1.50 .6361
SN/ML (°) 35.02 5.74 35.35 6.31 0.33 1.66 .0485
ANS-PNS/ML (°) 27.80 5.12 28.06 5.54 0.26 1.78 .1515
ANS-ME/N-ME (ratio) 58.21 2.34 58.54 2.35 0.32 1.29 .0147
Dentoalveolar
U1L/ANS-PNS (°) 110.38 5.56 108.05 5.39 ⫺2.33 6.62 .0006
U1 to A-Pog (mm) 9.03 2.41 6.30 1.86 ⫺2.72 1.88 ⬍.0001
L1L/ML (°) 97.92 6.33 103.18 6.57 5.26 4.94 ⬍.0001
L1 to A-Pog (mm) 1.66 1.99 3.82 1.86 2.15 1.63 ⬍.0001
Dental
Overbite (mm) 2.87 2.65 1.34 0.84 ⫺1.53 2.42 ⬍.0001
Overjet (mm) 7.28 1.84 2.48 0.85 ⫺4.81 1.80 ⬍.0001
Soft tissue
Nasolabial angle (°) 114.85 9.27 116.98 10.35 2.12 7.02 .0032
Ls to Sn-Pog’ (mm) 4.66 1.81 3.26 1.58 ⫺1.39 1.42 ⬍.0001
Li to Sn-Pog’ (mm) 3.44 1.98 2.63 1.91 ⫺0.80 1.46 ⬍.0001
Ls to E-line (mm) ⫺0.77 2.46 ⫺3.32 2.14 ⫺2.56 1.71 ⬍.0001
Li to E-line (mm) 0.10 2.33 ⫺1.55 2.31 ⫺1.64 1.67 ⬍.0001
N-No (mm) 48.62 3.77 51.54 4.07 2.92 2.15 ⬍.0001
Ls-U1 (mm) 11.01 2.32 13.47 2.04 2.47 1.59 ⬍.0001
Li-L1 (mm) 14.56 1.56 13.10 1.43 ⫺1.46 1.52 ⬍.0001

*Absolute difference.
†
Percentage improvement.

backward regression, only initial lip thickness and
change in overjet remained in the model. For every
millimeter of overjet reduction, a reduction of 0.396
mm of Ls to E-line is to be expected. For every
millimeter of upper lip thickness at T1, a 0.168-mm
reduction in Ls to E-line is expected. For the lower lip,
all variables except initial thickness and change in L1 to
A-Pog were excluded from the model. For every
millimeter of increase in L1 to A-Pog, an increase of
0.216 mm in Li to E-line is expected. For every
millimeter of lower lip thickness (Li-L1) at T1, a
reduction of 0.512 mm in Li to E-line is expected.
When we divided our pretreatment sample into 3
groups— horizontal (ANS-Me/N-Me ⱕ56%), normal
(56%⬍ ANS-Me/N-Me ⬍58%), and divergent (ANS-
Me/N-Me ⱖ58%)—we found no significant differ-
ences in the change of the measurements SN/ML (P ⫽
.429). The values for variable ANS-PNS/ML show that
the bite closed 0.1° in the divergent group and opened
0.8° and 0.9° in the horizontal and normal groups,
respectively. The difference between the vertical and
the other 2 groups was statistically significant (P ⫽
.018).
DISCUSSION
This study is the first to evaluate the quality of the
treatment results of Class II Division 1 patients treated
with extraction of the maxillary first molars. To do this,
we looked at the occlusion, and we described the
cephalometric and profile changes.
The PAR index was developed to quantify the
extent to which a dentition deviates from an ideally
formed dental arch and occlusion. It does not assess
other results, such as quality of life, reduced suscepti-
bility to oral diseases, and dysfunctions.13 The PAR
score gives a general impression of the dental arches
and the occlusion but does not take all dental variables
into account. For instance, in our study, the mandibular
incisors were slightly proclined at T2. This does not
have a negative effect on the PAR score. This is a
American Journal of Orthodontics and Dentofacial Orthopedics Stalpers et al 321
Volume 132, Number 3

Table III.
Prediction of soft-tissues changes, results of
backward regression analysis
Beta P value 95% CI for beta

Dependent variable
Ls to E-line (T2⫺T1)
Independent variables
Included
Constant 1.192 .245 ⫺0.831 3.216
Overjet (T2⫺T1) 0.396 .000 0.214 0.578
Ls-U1 at T1 ⫺0.168 .021 ⫺0.309 ⫺0.206
Excluded
U1 to A-Pog (T2⫺T1)
U1L/ANS-PNS (T2⫺T1)
Sex at T1
Fig 2. Nomogram with PAR score at T1 compared with Model fit Adj R2: .153
Dependent variable
PAR score at T2 and improvement categories; 73% of
Li to E-line (T2⫺T1)
patients were in “greatly improved” section; no patients Independent variables
were in “worse or no different” section. Included
Constant 5.349 .000 2.601 8.097
L1 to A-Pog (T2⫺T1) 0.216 .019 0.036 0.396
Li-L1 at T1 ⫺0.512 .000 ⫺0.701 ⫺0.323
Excluded
L1L/ML (T2⫺T1)
Overjet (T2⫺T1)
Sex at T1
Model fit Adj R2: .234

error. We chose to use the interobserver error because

Fig 3. Mean tracings: black, T1; red, T2. Note relatively
small changes in sagittal position of upper and lower
lips compared with greater changes in incisor position
when superimposed on palatal plane at ANS.
reason that we included cephalometric analysis in our
study.14
For all cephalometric variables and the PAR score,
a method error analysis was conducted, based on
duplicated measurements by a second orthodontist. In
most published studies, the intraobserver error is given.
This error is usually smaller than the interobserver
the intraobserver error cannot detect systematic errors
in the identification of cephalometric points. For 6 of
the cephalometric variables and the PAR score, a
systematic difference between the observers was found.
This was a matter of concern. However, for all 6
cephalometric variables, the systematic difference be-
tween the observers was small, always smaller than the
duplicate measurement error. The duplicate measure-
ment error in turn was much smaller than the standard
deviation of the variable concerned (a factor 4 for Li-L1
and more than a factor 10 for the other cephalometric
variables). In comparison with the variability of the
PAR scores in the patient group, the duplicate measure-
ment error for the PAR score was not as small (about
half of the standard deviation for the PAR score). But
when compared with the size of the treatment effect,
the duplicate measurement error was small (a differ-
ence of a factor 10). Therefore, the uncertainty added
by the measurement error was too small to consider
exclusion of variables from the analysis.
In this study, the average PAR score at T1 was high.
Eighty-four of the 100 patients had an initial score of
ⱖ22 points, and 15 was the lowest initial score.
Thirteen patients had a PAR score of zero at T2; 10 was
the highest PAR score at T2. Ten or fewer PAR points
are considered a slight deviation from normal occlu-
322 Stalpers et al
sion; 5 or fewer reflect nearly ideal occlusion.10,15
Previous studies showed that a 70% reduction of the
PAR score can be considered a great improvement in
occlusal factors.11,16 In our study, the sample showed
an 89.9% average reduction in PAR score. Other
studies ranged from a 68% reduction in PAR score in
orthodontically treated British patients to 78% in a
sample treated by Norwegian orthodontists to 88% in a
study of American Board of Orthodontics cases.17-19 In
our study, no patients were in the “worse or no
different” category at T2, whereas 73% of our patients
could be categorized as “greatly improved.” However,
all our patients had a Class II malocclusion with an
enlarged overjet. Because overjet has a weighting of 6
in the British weighting system, this adds considerably
to a high pretreatment PAR score. It is less difficult to
achieve striking, recordable changes when the initial
PAR score is high.20 Nevertheless, our study suggests
that the standard of the treatment was high.
For the soft-tissue profile changes during orthodon-
tic treatment, normal growth changes should be kept in
mind. It was shown that the upper and lower lips
become significantly more retruded to the E-line with
age.21,22 Ricketts23 found the lower lip in women to be
2.0 mm posterior to the E-line; it was slightly more
retruded in men. Bishara et al21 found the relative
position of the lower lip to be, on average, 1.7 mm
posterior to the E-line for adolescent boys and girls 15
years of age. In the sample of Paquette et al,5 the lower
lips retruded 3.1 mm in the 4-premolar extraction group
and 0.5 mm in the nonextraction group. In our study,
the lower lip at T1 was 0.1 mm anterior of the E-line.
At T2, the lower lip was 1.6 mm posterior of the E-line,
a retrusion of 1.8 mm. An explanation for this obser-
vation might be that, since all our patients had an
enlarged overjet, the lower lip rested on the protruded
maxillary incisors at T1. Overjet was, in our patients,
corrected by retraction of the maxillary incisors, pro-
clination of the mandibular incisors (since we did not
extract in the mandibular arch), and growth. When the
maxillary incisors are retracted, the lower lip follows
this movement. We were able to retract the maxillary
incisors with a relatively controlled retroclination of
2.3°, but we could not avoid proclination of the
mandibular incisors. The mandibular incisors pro-
clined on average 5.3° from 98.0° at T1 to 103.2° at
T2. This provided extra arch space and reduced the
amount of retraction needed for the maxillary inci-
sors. This in turn provided for an improved profile in
many patients.4 The standard deviations for the
changes in upper and lower lip thickness were high
(6.62 and 4.94, respectively). This indicates high
individual variability.
American Journal of Orthodontics and Dentofacial Orthopedics
September 2007
The variable upper lip position relative to the E-line
(Ls to E-line) was suggested by Ricketts23 as a way to
evaluate the position of the upper lip relative to the chin
and nose. He found that ideally the upper lip should be
4.0 mm posterior to this line in women and slightly
more retruded in men. These values were similar to the
adult values of Bishara et al.21 They found, for 15-year-
olds, that the upper lip was on average 2.9 mm posterior
of the E-line. We found, in our sample, the upper lip to
be on average 0.8 mm posterior of the E-line at T1 and
3.3 mm posterior of the E-line at T2. We retracted the
maxillary incisors on average 2.7 mm relative to the
A-Pog line and maintained good inclination. We ob-
served retrusion of the upper lip relative to the subna-
sale-soft-tissue-Pog line of 1.4 mm. This means that, in
our sample, the upper lip followed the movement of the
maxillary incisors for about 50% after extraction of the
first permanent maxillary molars. Also, Katsaros24
found only small soft-tissue profile changes compared
with the dental changes in identical twins, with 1 boy
treated with extraction of 4 premolars and the other
without extractions. The relatively small changes in the
sagittal position of the lips with A-Pog as a reference
line, compared with the larger changes measured with
the E-line as the reference, show that the influence of
the growth of the chin or the nose on the facial profile
might be more important than the orthodontic treatment.
In a stereophotogrammetric study on the growth of the
nose, early growth was found in girls and late growth in
boys, suggesting an adolescent growth spurt in the nose.25
Developmentally, the greatest change occurred in the
anteroposterior prominence of the nasal tip.
An important factor in our analysis was the predic-
tion of changes in lip position for the individual patient.
When we applied a backward regression model for the
upper lip, sex, change in maxillary incisor position, and
inclination did not seem to significantly influence the
change of the position of the upper lip in relation to the
E-line (Ls to E-line). However, change in overjet and
initial upper lip thickness could explain 15% of the
variation. For the prediction of the change of lower lip
position, sex, overjet, and inclination of the mandibu-
larr incisors did not seem to have significant influence.
However, mandibular incisor position (L1 to A-Pog)
and initial lower lip thickness (Li-L1) could explain
23% of the variation.
It was suggested that supraerupted maxillary first
molars are a consistent finding in patients with skeletal
open bite.5 Therefore, extraction of these molars might
aid in mandibular autorotation and thereby reduce
anterior facial height. In a sample of 8 patients with
anterior open bite treated orthodontically with the
extraction of 4 first molars, closing of the bite of 1.5°
American Journal of Orthodontics and Dentofacial Orthopedics
Volume 132, Number 3
(SN/GoGn angle) was found.26 These patients, how-
ever, had open bite extending to the posterior teeth. Our
findings are not consistent with this treatment philoso-
phy. The only statistically significant difference be-
tween T1 and T2 with regard to the skeletal vertical
measurements for the horizontal, normal, and divergent
groups was found for the measurement ANS-PNS/ML.
The bite closed 0.1° in the divergent group and opened
0.8° and 0.9° in the horizontal and normal groups,
respectively. This difference was statistically signifi-
cant, but the clinical relevance of this finding is limited.
In other words, we did not find a bite-closing effect of
our treatment. That means that divergent patients do not
seem to benefit more from this treatment modality than
do deepbite patients.
CONCLUSIONS
Our results suggest that extraction of the maxillary
first permanent molars in Class II Division 1 patients
leads to good treatment outcomes. Extraction of the
maxillary first permanent molars has only a minor
effect on the soft-tissue profile. The bite-closing effect
of extracting the maxillary first permanent molars is
statistically significant, but the clinical relevance is
questionable.
REFERENCES
1. Proffit WR, Fields HW, Moray LJ. Prevalence of malocclusion
and orthodontic treatment need in the United States: estimates
from the NHANES III survey. Int J Adult Orthod Orthognath
Surg 1998;13:97-106.
2. Steiner CC. Cephalometrics for you and me. Am J Orthod
1953;39:729-55.
3. Bennett JC, McLaughlin RP. Orthodontic management of the
dentition with the preadjusted appliance. Oxford: Isis Medical
Media; 1997. p. 231-50.
4. Subtelny JD, Sakuda M. Open bite diagnosis and treatment. Am J
Orthod 1964;50:337-58.
5. Paquette DE, Beattie JR, Johnston LE. A long-term comparison
of nonextraction and premolar extraction edgewise therapy in
“borderline” Class II cases. Am J Orthod Dentofacial Orthop
1992;102:1-14.
6. Katsaros C, Ripplinger B, Hoegel A, Berg R. The influence of
extraction versus non-extraction orthodontic treatment on the
soft tissue profile. J Orofac Orthop 1996;57:354-65.
7. Sandler PJ, Atkinson R, Murray AM. For four sixes. Am J
Orthod Dentofacial Orthop 2000;117:418-34.
8. Williams R. Single arch extraction— upper first molars or what to
do when nonextraction treatment fails. Am J Orthod 1979;76:
376-93.
Stalpers et al 323
9. Riolo ML, Moyers RE, McNamara JA Jr, Hunter WS. An atlas
of craniofacial growth: cephalometric standards from the Uni-
versity School growth study. Craniofacial Growth Series. Ann
Arbor: Center for Human Growth and Development; University
of Michigan; 1974.
10. Richmond S, Shaw WC, O’Brien KD, Buchanan IB, Jones R,
Stephens CD, et al. The development of the PAR index (peer
assessment rating: reliability and validity). Eur J Orthod 1992;
14:125-39.
11. Richmond S, Shaw WC, Roberts CT, Andrews M. The PAR
index (peer assessment rating): methods to determine the out-
come of orthodontic treatment in terms of improvement and
standards. Eur J Orthod 1992;14:180-7.
12. Dahlberg G. Statistical methods for medical and biological
students. New York: Interscience Publications; 1940.
13. Shaw WC, Richmond S, O’Brien KD. The use of occlusal
indices: a European perspective. Am J Orthod Dentofacial
Orthop 1995;107:1-10.
14. Birkeland K, Furevik J, Boe OE, Wist PJ. Evaluation of
treatment and posttreatment changes by the PAR index. Eur
J Orthod 1997;19:279-88.
15. Richmond S, Shaw WC, Stephens CD, Webb WG, Roberts CT,
Andrews M. Orthodontics in the general dental service of
England and Wales: a critical assessment of standards. Br Dent
J 1993;174:315-29.
16. DeGuzman L, Bahiraei D, Vig KW, Vig PS, Weyant MS,
O’Brien KD. The validation of the peer assessment rating index
for malocclusion severity and treatment difficulty. Am J Orthod
Dentofacial Orthop 1995;107:172-6.
17. O’Brien KD, Shaw WC, Roberts CT. The use of occlusal indices
in assessing the provision of orthodontic treatment by the
hospital orthodontic service of England and Wales. Br J Orthod
1993;20:25-35.
18. Richmond S, Andrews M. Orthodontic treatment standards in
Norway. Eur J Orthod 1993:15;7-15.
19. Dyken RA, Sadowsky PL, Hurst D. Orthodontic outcomes
assessment using the peer assessment rating index. Angle Orthod
2001;71:164-9.
20. Buchanan IB, Russell JI, Clark JD. An illustrative comparison of
the outcome of treatment using 2 fixed appliances techniques.
Br J Orthod 1996;23:351-7.
21. Bishara SE, Jakobsen JR, Hession TJ, Treder JE. Soft tissue
profile changes from 5 to 45 years of age. Am J Orthod
Dentofacial Orthop 1998;114:698-706.
22. Nanda RS, Meng H, Kapila S, Goorhuis J. Growth changes in the
soft tissue profile. Angle Orthod 1990;60:177-90.
23. Ricketts RM. Esthetics, environment, and the laws of lip relation.
Am J Orthod 1968;54:272-89.
24. Katsaros C. Profile changes following extraction vs. nonextrac-
tion orthodontic treatment in a pair of identical twins. J Orofac
Orthop 1996;57:56-9.
25. Burke PH, Hughes-Lawson CA. Stereophotogrammetric study of
growth and development of the nose. Am J Orthod Dentofacial
Orthop 1989;96:144-51.
26. Aras A. Vertical changes following orthodontic extraction treat-
ment in skeletal open bite subjects. Eur J Orthod 2002;24:407-16.