European Journal of Orthodontics, 2023, 45, 79–87

https://doi.org/10.1093/ejo/cjac035
Advance Access publication 6 July 2022
Original Article

Dentoskeletal and soft tissue changes after treatment of

crowding with premolar extractions: a 50-year follow-up
Nameer Al-Taai1, , Maurits Persson1, , Maria Ransjö1,2, , Eva Levring Jäghagen3, and
Anna Westerlund2,

Downloaded from https://academic.oup.com/ejo/article/45/1/79/6632281 by Umea University user on 06 March 2023

1
Orthodontics, Department of Odontology, Umeå University, Umeå, Sweden
2
Department of Orthodontics, Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden
3
Oral and Maxillofacial Radiology, Department of Odontology, Umeå University, Umeå, Sweden
Correspondence to: Nameer Al-Taai, Department of Odontology, Umeå University, SE-901 85 Umeå, Sweden. E-mail: nameer.al-taai@umu.se

Summary
Background: The long-term effects on dentofacial morphology of interceptive treatment with premolar extractions, in the absence of subse-
quent orthodontic treatment, have not been fully explored.
Objective: The aim was to investigate the effects of premolar extractions (without subsequent orthodontic treatment) on the dentoskeletal
and soft tissue profile of patients aged between 12 and 62 years with Class I malocclusion with severe crowding, as compared to untreated
controls.
Materials and methods: The Extraction group (N = 30 with Class I crowding malocclusion) had their first premolars removed in early adolescence
without subsequent orthodontic treatment. The Control group included 30 untreated subjects with Class I normal occlusion. Cephalograms were
taken at 12 (T1), 15 (T2), 30 (T3), and 62 (T4) years of age. A superimposition-based cephalometric method was used to assess the dentoskeletal
and soft tissue changes.
Results: There were no significant differences between the Extraction and Control groups in terms of skeletal sagittal relation, incisor inclin-
ation, and protrusion, or most of the soft tissue parameters throughout the observation period. However, significant differences were observed
between the groups with respect to the vertical relations in T2–T3, such that the Extraction group showed more-pronounced decreases in the
ML/NSL, ML/NL, and Gonial angles and more-pronounced increases in facial heights.
Conclusions and implications: Treatment for subjects with Class I malocclusion with severe crowding by the extraction of four premolars,
without subsequent orthodontic treatment, does not affect the long-term dentoskeletal and soft tissue profile, as compared to an untreated
Control group. The degree of crowding, rather than changes in dentofacial morphology, is crucial in deciding on extraction therapy.

Introduction and changes in the vertical facial dimension are considered to

Longitudinal growth studies have reported significant, be normal, age-related physiological changes (2,3).
age-related, craniofacial changes during adulthood (1–3). As a consequence of the current aesthetic ideal, whereby pa-
Interest in facial aesthetics has increased markedly in recent tients favour fuller lips, non-extraction orthodontic treatment
years. Cosmetic treatments, such as Botox and filler injections, options have become more popular. The potential changes in
are frequently used to counteract age-related facial changes lip position following premolar extraction and incisor retrac-
(4). Orthodontists also have the opportunity to influence the tion (7,8) may jeopardize the aesthetics. As a result, ortho-
facial profile by changing the soft tissue-supporting struc- dontists are reluctant to choose extraction treatment to relief
tures, e.g. the positions of the teeth and dental arches (5,6). crowding. Furthermore, the tendency to treat teeth crowding
Several studies have evaluated the effects of extractions and without extraction has been amplified by the marketing of
subsequent orthodontic treatment on craniofacial changes clear aligners and new bracket systems. This is the situation
(7–16). Some studies have reported no significant differences despite the fact that many cases should have been treated with
between the extraction and non-extraction orthodontic treat- extraction to achieve optimal outcomes.
ment outcomes (10,11,15,16). However, other studies have It can be argued that orthodontic treatment, following
found that premolar extraction can affect the dentoskeletal premolar extraction, may counteract adverse effects of the
and soft tissue profile (7–9,12). Possible reasons for these extraction on the soft tissue profile and vertical dimension.
discrepant results are differences in the study designs and It is, therefore, interesting to study the long-term effects on
inadequate control groups. Consequently, there is still no dentofacial structures of premolar extraction without subse-
clear consensus on the effects that tooth extraction has on quent orthodontic treatment. To date, only one study has in-
the dentoskeletal and soft tissue profile, despite the fact that vestigated the effect of serial extraction, without subsequent
crowding of the teeth is the most common type of malocclu- orthodontic treatment, on the soft tissue profile of patients
sion (17,18). Furthermore, continuous thinning of the lips aged up to 20 years (14).

© The Author(s) 2022. Published by Oxford University Press on behalf of the European Orthodontic Society
This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs licence (https://creativecommons.
org/licenses/by-nc-nd/4.0/), which permits non-commercial reproduction and distribution of the work, in any medium, provided the original work is not
altered or transformed in any way, and that the work is properly cited. For commercial re-use, please contact journals.permissions@oup.com
80 European Journal of Orthodontics, 2023

The aim of present study was to investigate the effects of

premolar extractions, without subsequent orthodontic treat-
ment, on the dentoskeletal and soft tissue changes in a group
of patients with Class I malocclusion with severe crowding
(Extraction group), from early adolescence to late adulthood,
compared with a matched control group of orthodontically
untreated subjects with Class I normal occlusion (Control
group).

Materials and methods

Design Figure 1. Flowchart of the subjects in the Extraction and Control groups,
listing the sex and mean age (SD) of the participants in the study and

Downloaded from https://academic.oup.com/ejo/article/45/1/79/6632281 by Umea University user on 06 March 2023

The study was approved by the Regional Ethical Board in indicating the numbers of dropouts in periods T1, T2, T3, and T4.
Umeå, Sweden (registration no. 2012-410-31M). Written in-
formed consent was obtained from all participants at 60 years
of age. The present study is a longitudinal study, in which these subjects had normal sagittal occlusion (without any
follow-ups are performed from early adolescence (12 years; malocclusions) with mean relative spaces of about +1 mm
T1) to late adolescence (15 years; T2) to early adulthood (30 and −1 mm in the upper and lower dental arches, respectively.
years; T3) and to late adulthood (62 years; T4). In addition, they had a normal soft tissue profile (23) and the
Lateral cephalograms were taken for all participants on all mean values of the overjet and overbite were 3 mm and 3 mm,
four occasions. For the controls, all the radiographic examin- respectively. The 30 untreated subjects were documented with
ations were justified for research purposes. In the Extraction cephalograms at mean ages of 12.8 years (T1), 15.7 years
group, the cephalometric examinations at T1 and T2 were (T2), and 30.8 years (T3), with one missing cephalogram at
justified as part of the treatment, whereas at T3 and T4 they T2. Only subjects still living in the public dental care region
were justified for research purposes. The radiation dose for were invited to participate in the study at T4. In total, 26
a lateral cephalogram is approximately 5.6 μSv, which is subjects were documented at a mean age of 61.6 years (T4),
equivalent to exposure to background radiation for 0.7 days although 4 subjects were excluded due to missing first molars
(19). This can be compared to a low-dose cone beam com- (Figure 1).
puted tomography (CBCT) examination, for which the radi- The exclusion criteria were: orthodontic treatment; max-
ation dose is 15- to 26-fold higher (20). illofacial surgery; using mandibular advancement devices for
treatment of snoring; craniofacial anomalies; missing teeth
Material and prosthodontic treatment including teeth mesial to the
The subjects in the study, when it was established in the upper and lower second molars.
1960s, were all patients in the Public Dental Health Care in
Umeå, Sweden (21,22).
Methods
The material in the present study included standardized
cephalograms from two groups. For both groups, the lateral cephalograms were analogue radio-
graphs, exposed using the same cephalostat, the Philips Super
Extraction group Rotalix x-ray tube (Philips, Germany), at T1, T2, and T3 with
magnification factor of 1.1 (in the midline). The cephalograms
The Extraction group consisted of 30 patients from Persson’s
at T4 were digital radiographs, acquired using the Cranex
study (21). Initially, these patients had Angle Class I crowding
cephalostat (Soredex, Helsinki, Finland), also with magnifica-
malocclusion, with a mean space deficiency of about 7 mm in
tion factor of 1.1 (in the midline). Linear measurements were
each dental arch (without other malocclusions) at 11 years
adjusted to a standardized enlargement of 10% (24).
of age (T1). In addition, they were classified as having a har-
All the cephalograms were acquired with habitual occlu-
monious soft tissue profile and the mean values of the overjet
sion and with relaxed lips. To conduct the digital cephalo-
and overbite were 3.9 mm and 3.8 mm, respectively. The 30
metric analysis, all the cephalograms from T1, T2, and T3
patients had all their first premolars extracted at a mean age
were scanned with the Epson Perfection V750 Pro digital
of 11.5 years, to relieve crowding, without subsequent ortho-
scanner (EPSON Europe B.V.) at a resolution of 250 dpi.
dontic treatment. More information on the consequences of
Thereafter, the cephalograms were imported as JPEG files
extraction spaces were presented in the study of Persson et al.
into the FACAD® ver. 3.9.2.1133 cephalometric software.
(21). The 30 patients were documented with cephalograms at
To enable calibration of the linear measurements in both
mean ages of 11.7 years (T1), 14.5 years (T2), and 30.5 years
groups, a calibration ruler was placed on each radiograph
(T3), with three and five missing cephalograms at T1 and T2,
during scanning at T1, T2, and T3, and the cephalostat
respectively. Only subjects still living in the public dental care
used at T4 was equipped with a calibration ruler. All the
region were invited to participate in the study at T4. In total,
cephalograms were coded in FACAD, so as to perform ran-
27 subjects were documented at a mean age of 62.8 years
domized tracing.
(T4), although 2 subjects were excluded due to missing first
molars (Figure 1).
Cephalometric analysis
Control group In all, 26 cephalometric landmarks and 6 lines were used
The Control group consisted of 30 untreated subjects from in the present study (Figure 2), and 42 angular and linear
the study of Thilander et al. (22). At 12 years of age (T1), cephalometric parameters were used to describe the dental,
N. Al-Taai et al. 81

The changes in 42 parameters, from T1 to T2, T2 to T3,

and T3 to T4 in the Extraction group, were compared with
the corresponding changes in the Control group (Tables 2–4).

Error of method
In order to evaluate the intra-observer reliability of the ceph-
alometric measurements, 20 randomly selected cephalograms
were retraced by the same orthodontist (NA-T) 3 months after
the initial tracing. The intra-observer reliability was assessed
by estimating the intra-class correlation coefficients (ICCs)
with 95% confidence intervals. The intra-observer reliability
of the cephalometric measurements was good, with ICCs in the

Downloaded from https://academic.oup.com/ejo/article/45/1/79/6632281 by Umea University user on 06 March 2023

range of 0.92–0.99 for most of the parameters. However, the
ICCs were 0.84 and 0.88 for the horizontal distance of menton
to vertical reference line (Me-NSLP) and nasolabial angle, re-
spectively. The inter-observer reliability of the cephalometric
measurements has been described by Al-Taai et al. (3).

Statistical analysis
Independent samples t-test were used to determine significant
differences in the skeletal, dental, and soft tissue patterns be-
tween the Extraction and Control groups at baseline (T1).
Tests of differences in changes to the dentoskeletal and soft
tissues between the Extraction and Control groups were car-
ried out using the independent samples t-test.
To maximize the use of available information in data,
the analyses of differences between time points have been
performed using a pairwise deletion approach rather than
listwise deletion. Thus, if a subject had measurements al-
lowing the calculation of a change between two time points,
this data point contributed to the analyses regardless if data
Figure 2. (a) Cephalometric landmarks. (b) Lines used in the present
were missing at some other time points for that subject.
study.
All statistical analyses were performed using the R ver.
4.0.0 software (R Core Team 2020), and the significance level
was set at P < 0.05.
skeletal, and soft tissue morphologies (Supplementary Table).
Two reference lines, the nasion-sella line (NSL) and a per-
pendicular line through the sella (NSLP), and two reference
Results
landmarks, sella (S) and nasion (N), were used in the T1 Cephalometric comparisons at baseline (T1) between the
cephalogram. Extraction and Control groups showed no significant differ-
A superimposition-based cephalometric method (25) was ences for the skeletal sagittal and vertical relations, incisor
used to measure 22 of 42 parameters at T2, T3, and T4 in re- inclination, and soft tissue profile (Table 1). Although signifi-
lation to the S and N landmarks at T1. Superimposition was cant differences were found between the two groups in the
performed on the anterior cranial base using the Tuberculum SNB, ANB, and facial profile angles, the values were within
Sella-Wing point method (25). After superimposition, ref- the normal range of orthognathic morphology (3).
erence lines NSL and NSLP and landmarks S and N were Tables 2–4 show the mean differences in the skeletal,
transferred digitally from the T1 cephalogram to the T2, T3, dental, and soft tissue changes, respectively, between the time
and T4 cephalograms. The FACAD program enabled meas- points for the Extraction and Control groups, and present the
urements of the 22 parameters for each of the T2, T3, and T4 tests of differences in the changes between the groups at the
cephalograms in relation to these transferred reference lines different periods.
and landmarks from the T1 cephalogram.
A conventional cephalometric method was used to perform Changes from T1 to T2
direct measurements (independent of the superimposition) of During the adolescence period, no significant differences in
the remaining 20 parameters. The cephalometric analysis and dentoskeletal and soft tissue changes were found between the
superimposition, for all included subjects, were performed groups, with two exceptions. The distance between the lower
digitally by one orthodontist (NA-T). incisor and A-Pogonion line (Ii-APog) increased 0.6 mm more
Similarities between the Extraction and Control groups in in the Extraction group, and the Mentolabial angle increased
terms of dentoskeletal and soft tissue morphologies at T1 in the Control group and decreased 5° in the Extraction
were required to exclude the influences of confounding fac- group (Tables 2–4).
tors on the studied parameters. Therefore, a comparison of
15 parameters, describing the dentoskeletal and soft tissue Changes from T2 to T3
patterns, was performed for the groups at baseline (T1) From late adolescence to early adulthood, the changes in
(Table 1). the sagittal position of the maxilla (SNA) and chin (SNPog)
82 European Journal of Orthodontics, 2023

Table 1. Mean values (SD) for the dental, skeletal and soft tissue parameters, comparing the Extraction and Control groups at baseline (T1).

Time point/parameters Extraction group (N = 27) Control group (N = 30) Test of differences between groups (P-value)
T1 T1

Sagittal relations
 SNA (°) 81.86 (3.51) 82.52 (3.16) 0.460
 SNB (°) 79.09 (3.49) 80.75 (2.47) 0.041
 ANB (°) 2.79 (1.72) 1.78 (1.57) 0.024
Vertical relations
 ML/NSL (°) 33.56 (4.95) 31.85 (4.45) 0.176
 NL/NSL (°) 6.21 (3.60) 6.00 (2.65) 0.800

Downloaded from https://academic.oup.com/ejo/article/45/1/79/6632281 by Umea University user on 06 March 2023

 ML/NL (°) 27.35 (4.73) 25.84 (4.55) 0.224
Dental relations
 ILs/NL (°) 109.87 (6.48) 111.81 (4.44) 0.191
 ILi/ML (°) 87.06 (5.92) 88.33 (4.69) 0.371
 Interincisal angle (°) 135.72 (9.77) 134.02 (7.20) 0.455
Soft tissue profile
 Total facial convexity (°) 142.80 (4.82) 144.22 (3.91) 0.227
 Facial profile angle (°) 166.06 (5.17) 168.95 (3.51) 0.016
 Nasolabial angle (°) 115.39 (10.99) 112.00 (8.61) 0.197
 Mentolabial angle (°) 134.96 (9.74) 136.07 (7.78) 0.634
 Upper lip thickness (mm) 11.98 (1.69) 11.99 (1.48) 0.983
 Lower lip thickness (mm) 13.30 (1.59) 12.63 (1.06) 0.066

Bold entries highlight the statistically significant value.

showed significantly more-pronounced (1°) forward growth with severe crowding has no effect on the long-term changes
in the Extraction group than in the Control group. Apart of the dentoskeletal and soft tissue profile, as compared with
from this, there were no significant differences in skeletal sa- an untreated control group classified as having normal oc-
gittal changes between the groups. clusal and dentofacial relations.
Significant differences were observed between the groups In the present study, a long-term (up to the sixth decade of life)
for changes in the skeletal vertical parameters. The man- assessment of dentoskeletal and soft tissue changes related to the
dibular inclination (ML/NSL), vertical jaw relation (ML/NL), extraction of premolars, without subsequent orthodontic treat-
and Gonial angle decreased to greater extents (1.7°, 1.6°, ments, was performed. Given that the Extraction and Control
and 2.2°, respectively) in the Extraction group. Moreover, groups showed similar dentoskeletal and soft tissue morpholo-
the anterior, lower anterior, posterior, and lower posterior gies at baseline, the groups were considered to be well-matched.
facial heights increased to greater extents (2 mm, 1.5 mm, Thus, potentially confounding effects related to sample selection,
3.8 mm, and 2.8 mm, respectively) in the Extraction group. orthodontic treatment, and/or growth were avoided in this study.
No statistically significant differences were found between As previously reported (21), the residual extraction spaces
the groups in terms of changes to the incisor inclination. closed in almost all the cases already at T3, and this closure
Four of the twelve soft tissue profile parameters showed was rapid initially and achieved by tipping of the adjacent teeth
significant differences between the groups. The anterior (MEs- in the lower arch. Spontaneous alignment and space closure in
NSL) and lower anterior (NL-MEs) soft tissue facial heights Class I crowding cases, following the extraction of four pre-
increased more, by 3 mm and 2 mm, respectively, in the molars, may be part of a so-called ‘serial extraction treatment’.
Extraction group. Furthermore, the protrusions of the upper To date, only one study has investigated the effect of serial ex-
and lower lips in relation to the vertical reference line (Ls-NSLP tractions, without subsequent orthodontic treatment, on the
and Li-NSLP) increased in the Extraction group (Tables 2–4). soft tissue profile from 13 to 24 years of age (14). In the 10-year
follow-up (14), the serial extraction group was compared with
Changes from T3 to T4 patients who underwent serial extraction and orthodontic treat-
From early to late adulthood, the groups exhibited no signifi- ment and other patients who underwent late premolar extrac-
cant differences in 39 of the 42 parameters. The skeletal vertical tion and orthodontic treatment. No significant differences were
parameters, ML/NSL and ML/NL angles, increased more in the found in relation to the changes in the soft tissue profile be-
Control group, and the distance from the upper lip to the aesthetic tween these three different treatment methods (14). However,
line (Ls-EL) increased more in the Extraction group (Tables 2–4). that study did not include an untreated control group, and the
follow-up period was shorter (14), making comparison with the
present study difficult. This means that with the present study
Discussion we add 40 years of knowledge to this topic.
The current study shows that early extraction of four pre- In the present study, the groups showed similar changes
molars as treatment for patients with Class I malocclusion in the skeletal sagittal relations throughout the observation
 N. Al-Taai et al.

Table 2. Differences in the mean values (SD) of the skeletal changes between time points for the Extraction and Control groups, together with tests of the differences in the changes between the groups at
different periods.

Periods/parameters Extraction group Control group Test of differences between groups

T1–T2 (N = 25) T2–T3 (N = 25) T3–T4 (N = 25) T1–T2 (N = 29) T2–T3 (N = 29) T3–T4 (N = 22) P-value P-value P-value
T1 versus T2 T2 versus T3 T3 versus T4

Sagittal relations
 SNA (°)* 2.17 (1.90) 1.63 (2.33) −0.86 (1.22) 2.33 (2.04) 0.49 (1.75) −0.97 (1.52) 0.759 0.046 0.795
 SNB (°)* 1.37 (1.85) 1.03 (1.95) −0.94 (1.30) 1.36 (1.56) −0.05 (2.04) −1.16 (1.27) 0.978 0.054 0.569
 ANB (°)* 0.78 (1.21) 0.60 (1.45) 0.08 (0.81) 0.95 (1.07) 0.57 (1.03) 0.16 (0.82) 0.581 0.927 0.726
 SNPog (°)* 1.69 (1.99) 1.26 (1.99) −0.82 (1.47) 1.40 (1.58) 0.09 (2.08) −1.16 (1.25) 0.551 0.039 0.397
 A-NSLP (mm)* 1.51 (1.83) 1.34 (2.31) −0.89 (1.37) 1.82 (1.88) 0.31 (1.81) −0.96 (1.56) 0.545 0.074 0.876
 B-NSLP (mm)* 1.11 (2.87) 0.97 (3.09) −1.44 (2.39) 1.33 (2.43) −0.42 (3.40) −1.90 (2.22) 0.759 0.125 0.496
 Pog-NSLP (mm)* 1.80 (3.34) 1.49 (3.59) −1.50 (3.04) 1.68 (2.80) −0.16 (3.96) −2.15 (2.50) 0.889 0.117 0.432
 Me-NSLP (mm)* 1.37 (3.57) 1.22 (3.80) −1.42 (3.36) 1.54 (2.66) −0.35 (4.69) −2.38 (2.59) 0.839 0.188 0.283
Vertical relations
 ML/NSL (°)* −0.60 (2.32) −1.34 (2.71) 1.05 (1.69) −0.73 (1.51) 0.40 (2.30) 2.08 (1.36) 0.798 0.013 0.027
 NL/NSL (°)* 0.89 (2.33) 0.30 (1.83) 1.05 (1.68) 0.64 (1.44) 0.57 (2.06) 0.31 (1.85) 0.636 0.616 0.158
 ML/NL (°) −1.31 (2.27) −1.81 (3.01) 0.28 (1.55) −1.37 (1.22) −0.14 (1.75) 1.75 (1.52) 0.913 0.014 0.002
 N-Me (mm)* 6.68 (3.33) 6.25 (5.07) −0.05 (1.86) 6.02 (3.14) 4.14 (2.11) 0.23 (1.66) 0.453 0.047 0.592
 ANSʹ-Me (mm) 3.65 (1.81) 3.99 (3.32) −0.65 (1.44) 3.72 (1.93) 2.59 (1.74) 0.15 (1.27) 0.888 0.053 0.053
 S-Go (mm)* 5.70 (2.72) 6.56 (5.25) −1.14 (1.49) 5.43 (2.62) 2.77 (2.75) −2.01 (1.45) 0.713 0.001 0.050
 PNSʹ-Go (mm) 3.02 (2.45) 4.55 (3.78) −0.87 (1.67) 3.57 (1.76) 1.75 (2.33) −1.71 (1.54) 0.348 0.002 0.079
 A-NSL (mm)* 4.15 (2.52) 3.37 (3.30) 0.18 (1.43) 3.01 (1.92) 2.18 (1.45) −0.40 (1.67) 0.065 0.084 0.205
 B-NSL (mm)* 5.59 (2.71) 4.37 (4.08) −0.70 (1.80) 5.14 (2.32) 2.95 (1.48) −0.26 (1.72) 0.513 0.087 0.397
 Pog-NSL (mm)* 6.82 (3.27) 6.21 (5.08) −0.37 (1.65) 5.95 (3.13) 4.01 (2.07) −0.55 (1.89) 0.326 0.037 0.726
 Me-NSL (mm)* 7.20 (3.27) 6.54 (5.38) −0.33 (1.57) 6.42 (3.38) 4.03 (2.11) −0.20 (1.44) 0.400 0.024 0.774
 Ar/Go/Me (°) −2.03 (1.96) −3.62 (2.48) 0.85 (1.53) −2.20 (1.67) −1.35 (2.27) 1.18 (2.34) 0.724 <0.001 0.566

Bold entries highlight the statistically significant value.

*Measured according to superimposition-based cephalometric method (25).
83

Downloaded from https://academic.oup.com/ejo/article/45/1/79/6632281 by Umea University user on 06 March 2023

84 European Journal of Orthodontics, 2023

period. Similar results have been obtained in short-term

P-value
Test of differences between groups

versus

0.641
0.310
0.843
0.912
0.588
0.177
0.577
0.793
studies of orthodontically treated patients (26,27). However,

T3

T4
the significant differences noted in the present study between
Table 3. Differences in the mean values (SD) of the dental relation changes between time points for the Extraction and Control groups, together with tests of the differences in the changes between the

the two groups with respect to the changes in the SNA and
SNPog angles, from late adolescence to early adulthood, in-
P-value

versus dicate slightly more anteriorly directed growth of the max-

0.657
0.078
0.086
0.880
0.547
0.075
0.630
0.319
illa and chin in the Extraction group. We consider that this
T2

T3
difference is not a result of the premolar extractions. Thus,
in clinical terms, premolar extractions in cases of Class I mal-
occlusion with severe crowding, without subsequent ortho-
P-value

0.399
0.395
<0.001
0.397
0.089
0.453
0.425
0.132
versus

dontic treatment, do not affect the skeletal sagittal growth

T1

T2

pattern.
It has been reported that changes in the inclination of

Downloaded from https://academic.oup.com/ejo/article/45/1/79/6632281 by Umea University user on 06 March 2023

the upper and lower incisors induced by orthodontic treat-
T3–T4 (N = 22)

ment may result in positional changes of the A and B land-

0.14 (1.32)
−0.09 (1.02)
−1.56 (2.14)
−0.14 (0.72)
3.88 (3.94)
−1.13 (1.79)
−3.56 (3.15)
−2.08 (2.53)

marks, which in turn result in a change to the sagittal jaw

relation (ANB) (28,29). It is noteworthy that throughout
the entire observation period, we found no significant dif-
ferences between the groups for the eight studied parameters
that describe the inclination and protrusion of the incisors.
The incisor retroclination observed in the Extraction group
Control group

T2–T3 (N = 29)

throughout the observation period was similar to that in the

1.15 (1.16)
−0.20 (2.85)
−0.13 (0.68)
1.12 (1.17)
0.14 (4.57)
0.12 (2.08)
−0.89 (3.35)
0.91 (3.45)

Control group and can be attributed to age-related changes.

Moreover, incisor relations expressed by overjet and over-
bite exhibited no significant differences between the groups
throughout the observation period (30). Therefore, in the
clinical context, premolar extractions in patients with Class
I malocclusion with severe crowding, without subsequent
T1–T2 (N = 29)

orthodontic treatment, do not affect the inclination and pro-

trusion of the upper and lower incisors. This is in contrast
1.29 (0.77)
1.99 (1.21)
1.49 (2.08)
−0.18 (0.61)
1.26 (3.28)
1.65 (1.96)
−0.66 (2.72)
0.75 (2.23)

to the results obtained in short-term studies, showing greater

retroclination of the incisors in orthodontically treated pa-
tients with extractions, as compared to non-extraction treat-
ments (9,26,27).
The groups showed similar skeletal vertical changes during
the adolescence period. This accords with the results of short-
T3–T4 (N = 25)

term studies that compared the effects of orthodontic treat-

−0.26 (1.38)
−0.03 (1.00)
−1.21 (2.21)
−0.24 (0.78)
3.53 (5.10)
−1.08 (1.64)
−2.24 (3.38)
−1.58 (3.45)

ment with and without extractions (15,16,27). Therefore,

the concept that premolar extraction, without subsequent
orthodontic treatment, affects facial heights or vertical jaw
relations is based on a low level of evidence, at least during
*Measured according to superimposition-based cephalometric method (25).

adolescence.
Extraction group

However, from late adolescence to early adulthood, the

T2–T3 (N = 25)

skeletal facial heights (anterior, lower anterior, posterior, and

1.30 (1.70)
1.88 (1.82)
1.15 (2.81)
−0.09 (1.17)
1.11 (7.06)
1.31 (2.71)
−1.34 (3.53)
2.02 (4.64)

lower posterior) increased to greater extents in the Extraction

group than in the Control group. This might be explained by
tipping and extrusion of the posterior teeth concomitant with
Bold entries highlight the statistically significant value.

mesial migration, following premolar extraction, as observed

earlier (21).
In contrast to the similar changes observed for the maxil-
T1–T2 (N = 25)

lary inclination (NL/NSL) in the two groups throughout the

1.54 (1.35)
1.72 (1.14)
0.96 (2.43)
−0.81 (0.64)
3.18 (4.84)
1.24 (2.08)
−1.30 (3.17)
−0.54 (3.86)

observation period, the changes in mandibular inclination

(ML/NSL) differed significantly between the groups from
late adolescence to early adulthood and from early to late
adulthood.
groups at different periods.

The extraction group exhibited a slight anterior mandibular

rotation and slightly greater degree of deepening of the ver-
Periods/parameters

tical jaw relation (ML/NL) and Gonial angle, as compared to

Isl-NSLP (mm)*
Iil-NSLP (mm)*

the Control group, from late adolescence to early adulthood.

Isa-Is/Iia-Ii (°)

Ii-APog (mm)
Isa-Is/NL (°)

This is attributed to the larger increase in the posterior facial

Iia-Ii/ML (°)

Ii-ML (mm)
Is-NL (mm)

height, about 4 mm, following premolar extraction.

From early to late adulthood, the Extraction group showed
lower degrees of posterior mandibular rotation (1°) and
 N. Al-Taai et al.

Table 4. Differences in the mean values (SD) of the soft tissue profile changes between time points for the Extraction and Control groups, together with tests of the differences in the changes between the
groups at different periods.

Periods/parameters Extraction group Control group Test of differences between groups

T1–T2 (N = 25) T2–T3 (N = 25) T3–T4 (N = 25) T1–T2 (N = 29) T2–T3 (N = 29) T3–T4 (N = 22) P-value P-value P-value
T1 versus T2 T2 versus T3 T3 versus T4

GL/PRN/PGs (°) −2.26 (2.63) −0.84 (3.10) 0.67 (2.51) −1.41 (3.26) −0.71 (2.91) 0.37 (2.14) 0.301 0.871 0.664
GL/SN/PGs (°) 0.24 (2.37) 1.96 (3.26) 5.13 (3.14) 0.08 (2.34) 2.39 (2.80) 4.07 (2.49) 0.803 0.608 0.209
MEs-NSL (mm)* 7.99 (3.46) 7.35 (6.53) 0.63 (2.23) 6.83 (3.57) 4.03 (2.43) 0.93 (1.78) 0.235 0.014 0.617
NL-MEs (mm) 4.48 (2.25) 4.86 (4.49) 0.08 (2.08) 4.18 (2.18) 2.65 (2.12) 1.11 (1.77) 0.621 0.021 0.075
MS/SN/Ls (°) −0.64 (7.85) −2.01 (6.73) −3.42 (8.30) 0.38 (5.65) −2.10 (7.27) −5.49 (6.82) 0.583 0.962 0.359
Li/Sli/PGs (°) −5.18 (6.41) 7.48 (9.04) 3.74 (9.03) 0.40 (6.48) 4.51 (6.02) 6.18 (9.58) 0.003 0.156 0.373
Ls-EL (mm) −1.77 (1.24) −2.75 (1.91) −1.86 (1.32) −1.52 (1.53) −2.61 (1.54) −0.71 (1.95) 0.509 0.759 0.022
Li-EL (mm) −1.29 (1.29) −2.14 (1.73) −1.88 (1.43) −0.98 (1.29) −1.81 (1.58) −1.37 (1.43) 0.379 0.468 0.223
Isl-Ls (mm) 1.10 (1.27) −0.74 (1.76) −1.61 (1.24) 0.73 (1.54) −1.46 (1.41) −0.92 (1.80) 0.351 0.102 0.130
Iil-Li (mm) 1.12 (1.16) 0.28 (1.68) −0.68 (1.36) 0.88 (1.18) −0.16 (0.97) −0.63 (1.34) 0.462 0.238 0.886
Ls-NSLP (mm)* 2.41 (2.74) 0.37 (3.74) −3.55 (2.27) 2.36 (2.45) −1.57 (2.19) −3.00 (2.56) 0.936 0.022 0.441
Li-NSLP (mm)* 2.00 (3.10) 1.64 (3.64) −2.22 (2.57) 2.56 (2.62) −0.53 (2.68) −2.50 (2.84) 0.477 0.015 0.724
Ls-NSL (mm)* 4.56 (3.15) 5.26 (4.43) 3.61 (2.33) 3.95 (2.10) 3.98 (2.19) 3.96 (2.68) 0.402 0.175 0.633
Li-NSL (mm)* 5.91(3.80) 3.75(3.92) 0.94(2.74) 4.27(2.57) 3.05(2.61) 1.20(3.21) 0.066 0.441 0.759

Bold entries highlight the statistically significant value.

*Measured according to superimposition-based cephalometric method (25).
85

Downloaded from https://academic.oup.com/ejo/article/45/1/79/6632281 by Umea University user on 06 March 2023

86 European Journal of Orthodontics, 2023

opening of the ML/NL angle (1.5°), as compared to the Clinical implications

Control group. This reflected slight differences in the changes In cases of Class I malocclusion with severe crowding, the
of the anterior relative to the posterior facial heights between degree of crowding, rather than future changes in soft tissue
the groups. Given these small differences between the groups, profile and incisor inclination, should be the main consider-
we cannot attribute them to premolar extraction. However, ation when choosing an extraction therapy in the absence of
from early to late adulthood, both groups showed low levels subsequent orthodontic treatment.
of posterior mandibular rotation and opening of the vertical
jaw relation. This is due to the remodelling that occurs within Limitations
the region of the angle of the mandible from early to late Assessment of the growth-related changes were performed on
adulthood, as observed by Al-Taai and colleagues (3). 2D radiographs rather than 3D (CBCT) images. This technique
Clinically, the extraction of premolars in patients with was not available 50 years ago and is questionable from the
Class I malocclusion with severe crowding, without subse- radiation protection and ethical perspectives. The radiation
quent orthodontic treatment, may increase the facial heights

Downloaded from https://academic.oup.com/ejo/article/45/1/79/6632281 by Umea University user on 06 March 2023

dose from a lateral cephalogram is low, approximately 5.6
and may, simultaneously, result in slight anterior mandibular μSv, which is equivalent to 0.7 days of average background
rotation and deepening of the vertical jaw relation from late radiation to which general population in, for example, the
adolescence to early adulthood. Thereafter, from early to late USA or Sweden is exposed (19). This can be compared to a
adulthood, premolar extraction results in relatively stable fa- low-dose CBCT examination, where the dose is 15–26 times
cial heights concomitant with slight posterior mandibular ro- higher (20). Therefore, the widely used 2D lateral cephalo-
tation and opening of the vertical jaw relationship. metric radiographs can be justified from the radiation protec-
Both groups showed similar changes to the soft tissue tion perspective, and are still useful when long-term assessment
profile, with the exceptions of the changes in the Mentolabial of craniofacial changes is performed based on superimposition.
angle and position of the upper lip during the adolescence In the present study, cephalometric measurements of 22 linear
period and from early to late adulthood, respectively. The and angular parameters at T2, T3, and T4 were performed using
decrease in the Mentolabial angle in the Extraction group a superimposition-based method (25). Thus, the age-related
might be due to soft tissue growth rather than the extrac- positional changes in the nasion and sella landmarks were ex-
tion itself, given that similar changes in the sagittal relations cluded when assessing parameters related to the cranial base.
(SNB, SNPog, B-NSLP, and Pog-NSLP) and dental inclin- In addition, the cephalograms acquired at T1, T2, and T3
ations (Iia-Ii/ML and Iil-NSLP) were noted in the two groups. needed to be scanned. The scanning process may have caused
The more-retruded upper lip in relation to the aesthetic line, slightly reduced resolution and some blurring.
as seen in the Extraction group, may reflect more-pronounced In the current study, there were more women than men
thinning of the upper lip, since similar changes in the sagittal in the Control group. Given the previous findings (1,2) that
relations (SNA and A-NSLP) and dental inclination (Isa-Is/ men and women exhibit different directions of mandibular
NL and Isl-NSLP) were observed in the two groups. Short- rotation, this may have skewed the comparison with the
term and long-term studies have reported more-retruded lips Extraction group. However, when the groups were merged
in patients who underwent premolar extractions and ortho- (to achieve higher power) we observed that men showed the
dontic treatment (9,11,26,27). same vertical jaw inclination as women during T1–T2 and
From late adolescence to early adulthood, the Extraction T2–T3 in this material.
group showed more-pronounced increases in the anterior and Finally, it could be argued that the pronounced increases
lower anterior soft tissue facial heights than did the Control in facial heights in the Extraction group are a consequence of
group, probably due to the more-pronounced skeletal vertical more women being included in the Control group, since it has
growth observed in the Extraction group. In addition, the been observed (2) that men record higher values for the linear
slight protrusion of the lips in the Extraction group compared craniofacial dimensions than women. Nevertheless, when we
to the Control group from late adolescence to early adult- performed a gender-based analysis, the men and women, re-
hood likely reflects a more anteriorly directed growth of the spectively, in the Extraction group showed greater increases
jaws rather than an effect of the extraction itself. in the facial heights than those in the Control group.
Among the soft tissue parameters that exhibited signifi-
cant differences between groups, the differences were small Conclusions
and could, therefore, not be considered clinically significant.
Consequently, the extraction of four premolars in cases of Treatment for subjects with Class I malocclusion with severe
Class I malocclusion with severe crowding does not affect the crowding by the extraction of four premolars, without subse-
long-term soft tissue profile changes, which include: the facial quent orthodontic treatment, does not affect, in general, the
profile angle, total facial convexity or nasolabial angle, and long-term changes in the skeletal relations, incisor inclination
lip thickness or length, as well as the lower lip distance to the or protrusion, and lip support or soft tissue profile, as com-
aesthetic line. This is consistent with the results of a study that pared to untreated controls.
compared the soft tissue profile changes in serial extraction Therefore, in Class I malocclusion with severe crowding, the
and late premolar extraction treatments from 13 to 24 years degree of crowding, rather than changes to the dentoskeletal
of age (14), and also with long-term follow-up studies of ex- and facial aesthetics, is the crucial factor in deciding to employ
tractions with subsequent orthodontic treatment (11,13). extraction therapy without subsequent orthodontic treatment.
It has been reported in short-term studies that lip pos-
ition can be affected by the incisor position after extraction
(9,26,27). However, our study could not confirm this, since
Acknowledgements
both groups showed similar changes in incisor inclination and We thank Per Liv, PhD, certified statistician for valuable sup-
lip position. port with the statistical analyses.
N. Al-Taai et al. 87

Funding nonextraction patients. American Journal of Orthodontics and

Dentofacial Orthopedics, 128, 450–457.
This study was supported financially by the Medical Faculty 14. Wilson, J.R., Little, R.M., Joondeph, D.R. and Doppel, D.M. (1999)
of Umeå University, Sweden. Comparison of soft tissue profile changes in serial extraction and
late premolar extraction. The Angle Orthodontist, 69, 165–173;
discussion 73–74.
Conflicts of interest 15. Kocadereli, I. (1999) The effect of first premolar extraction on ver-
None to declare. tical dimension. American Journal of Orthodontics and Dentofacial
Orthopedics, 116, 41–45.
16. Staggers, J.A. (1994) Vertical changes following first premolar
Data availability extractions. American Journal of Orthodontics and Dentofacial
The data that form the basis for this article are presented in Orthopedics, 105, 19–24.
the paper. 17. Josefsson, E., Bjerklin, K. and Lindsten, R. (2010) Self-perceived
orthodontic treatment need and prevalence of malocclusion in

Downloaded from https://academic.oup.com/ejo/article/45/1/79/6632281 by Umea University user on 06 March 2023

18- and 19-year-olds in Sweden with different geographic origin.
Supplementary material
Swedish Dental Journal, 34, 95–106.
Supplementary material is available at European Journal of 18. Proffit, W.R., Fields, H.W., Jr. and Moray, L.J. (1998) Prevalence of
Orthodontics online. malocclusion and orthodontic treatment need in the United States:
estimates from the NHANES III survey. International Journal of
Adult Orthodontics and Orthognathic Surgery, 13, 97–106.
References 19. Ludlow, J.B., Davies-Ludlow, L.E. and White, S.C. (2008) Patient
1. Pecora, N.G., Baccetti, T. and McNamara, J.A., Jr. (2008) The aging risk related to common dental radiographic examinations: the im-
craniofacial complex: a longitudinal cephalometric study from late pact of 2007 International Commission on Radiological Protection
adolescence to late adulthood. American Journal of Orthodontics recommendations regarding dose calculation. Journal of the Amer-
and Dentofacial Orthopedics, 134, 496–505. ican Dental Association, 139, 1237–1243.
2. Behrents, R.G. (1985) Growth in the Aging Craniofacial Skeleton. 20. Signorelli, L., Patcas, R., Peltomäki, T. and Schätzle, M. (2016)
Monograph 17. Craniofacial Growth Series. Center for Human Radiation dose of cone-beam computed tomography compared
Growth and Development; University of Michigan, Ann Arbor, MI. to conventional radiographs in orthodontics. Journal of Orofacial
3. Al-Taai, N., Persson, M., Ransjö, M., Levring Jäghagen, E., Fors, Orthopedics, 77, 9–15.
R. and Westerlund, A. (2022) Craniofacial changes from 13 to 62 21. Persson, M., Persson, E.C. and Skagius, S. (1989) Long-term sponta-
years of age. European Journal of Orthodontics, XX, 1–10. neous changes following removal of all first premolars in Class I cases
4. Wise, J.B. and Greco, T. (2006) Injectable treatments for the aging with crowding. European Journal of Orthodontics, 11, 271–282.
face. Facial Plastic Surgery, 22, 140–146. 22. Thilander, B., Persson, M. and Skagius, S. (1982)
5. Pancherz, H., Bjerklin, K., Lindskog-Stokland, B. and Hansen, K. Roentgencephalometric standards for the facial skeleton and soft
(2014) Thirty-two-year follow-up study of Herbst therapy: a bio- tissue profile of Swedish children and young adults. II. Comparisons
metric dental cast analysis. American Journal of Orthodontics and with earlier Scandinavian normative data. Swedish Dental Journal.
Dentofacial Orthopedics, 145, 15–27. Supplement, 15, 219–228.
6. Masucci, C., Franchi, L., Defraia, E., Mucedero, M., Cozza, P. 23. Proffit, W.R., Fields, H. W., Larson, B. E. and Sarver, D. M. (2019)
and Baccetti, T. (2011) Stability of rapid maxillary expansion and Contemporary Orthodontics. Mosby Elsevier, St. Louis, MO.
facemask therapy: a long-term controlled study. American Journal 24. Hatton, M.E. and Grainger, R.M. (1958) Reliability of
of Orthodontics and Dentofacial Orthopedics, 140, 493–500. measurements from cephalograms at the Burlington Orthodontic
7. Akyalçin, S., Hazar, S., Güneri, P., Gögüs, S. and Erdinç, A.M. (2007) Research Centre. Journal of Dental Research, 37, 853–859.
Extraction versus non-extraction: evaluation by digital subtraction 25. Al-Taai, N., Levring Jäghagen, E., Persson, M., Ransjö, M. and
radiography. European Journal of Orthodontics, 29, 639–647. Westerlund, A. (2021) A superimposition-based cephalometric
8. Bishara, S.E., Cummins, D.M., Jakobsen, J.R. and Zaher, A.R. method to quantitate craniofacial changes. International Journal of
(1995) Dentofacial and soft tissue changes in Class II, division 1 Environmental Research and Public Health, 18, 5260.
cases treated with and without extractions. American Journal of 26. Germeç, D. and Taner, T.U. (2008) Effects of extraction and
Orthodontics and Dentofacial Orthopedics, 107, 28–37. nonextraction therapy with air-rotor stripping on facial esthetics
9. Kocadereli, I. (2002) Changes in soft tissue profile after ortho- in postadolescent borderline patients. American Journal of Ortho-
dontic treatment with and without extractions. American Journal dontics and Dentofacial Orthopedics, 133, 539–549.
of Orthodontics and Dentofacial Orthopedics, 122, 67–72. 27. Kirschneck, C., Proff, P., Reicheneder, C. and Lippold, C. (2016)
10. Kouvelis, G., Dritsas, K., Doulis, I., Kloukos, D. and Gkantidis, N. Short-term effects of systematic premolar extraction on lip profile,
(2018) Effect of orthodontic treatment with 4 premolar extractions vertical dimension and cephalometric parameters in borderline pa-
compared with nonextraction treatment on the vertical dimension tients for extraction therapy—a retrospective cohort study. Clinical
of the face: a systematic review. American Journal of Orthodontics Oral Investigations, 20, 865–874.
and Dentofacial Orthopedics, 154, 175–187. 28. Al-Abdwani, R., Moles, D.R. and Noar, J.H. (2009) Change of inci-
11. Rathod, A.B., Araujo, E., Vaden, J.L., Behrents, R.G. and Oliver, sor inclination effects on points A and B. The Angle Orthodontist,
D.R. (2015) Extraction vs no treatment: long-term facial profile 79, 462–467.
changes. American Journal of Orthodontics and Dentofacial Or- 29. Sun, Q., Lu, W., Zhang, Y., Peng, L., Chen, S. and Han, B. (2021)
thopedics, 147, 596–603. Morphological changes of the anterior alveolar bone due to retrac-
12. Sivakumar, A. and Valiathan, A. (2008) Cephalometric assessment tion of anterior teeth: a retrospective study. Head & Face Medicine,
of dentofacial vertical changes in Class I subjects treated with 17, 30.
and without extraction. American Journal of Orthodontics and 30. Persson, M., Al-Taai, N., Pihlgren, K. and Westerlund, A. (2022)
Dentofacial Orthopedics, 133, 869–875. Early extractions of premolars reduce age-related crowding of
13. Stephens, C.K., Boley, J.C., Behrents, R.G., Alexander, R.G. and lower incisors: 50 years of follow-up. Clinical Oral Investigations,
Buschang, P.H. (2005) Long-term profile changes in extraction and 26, 4525–4535.