Original Article

Airway and cephalometric changes in adult orthodontic patients after

premolar extractions
Adrienne Joya; Joorok Parkb; David William Chambersc; Heesoo Ohd

ABSTRACT
Objective: To examine changes in the airway and cephalometric measurements associated with
orthodontic treatment of adults with and without premolar extractions. The study investigated
whether extractions had a direct or indirect effect on the airway and examined selected skeletal and
dental features.
Materials and Methods: This retrospective study used pre- (T1) and posttreatment (T2) cone-
beam computed tomography scans of 83 adult patients matched for age and sex. A total of 15
airway and 10 skeletal and dental measures were analyzed by means of repeated-measures
analysis of variance.
Results: There were no results showing that extractions affected airway dimensions that could not
be accounted for as reflections of measurement error. There was no evidence that extractions
affected the airway indirectly through skeletal and dental changes. There were strong and consistent
findings that patients with small airways showed larger ones after treatment and that patients with
large airways showed smaller ones later. These effects were independent of whether or not
extractions were part of treatment. The measurement phenomena of regression toward the mean and
of differential unfolding of natural changes over time could have accounted for the results observed.
Conclusions: There was no evidence that extractions in nongrowing patients have negative
consequences on the size of various airway measures in the nasopharynx, retropalatal, or
retroglossal regions. (Angle Orthod. 2020;90:39–46.)
KEY WORDS: Airway change; Premolar extraction; Adults; CBCT; MCA; Incisor retraction

INTRODUCTION beam computed tomography (CBCT) and concerns

regarding obstructive sleep apnea.1–3 In addition, it
Increased attention is being paid to airway capacity,
seems there is a trend toward less frequent extraction
promoted in part by more widespread use of cone-
and more arch expansion, especially in the mixed
dentition.4 These trends are sometimes conjoined and
a
Adjunct Clinical Instructor, Department of Orthodontics,
University of the Pacific, Arthur A. Dugoni School of Dentistry,
supported by the belief that extraction of premolars
San Francisco, Calif. contributes to collapse of the airway. In a survey of 125
b
Assistant Professor, Department of Orthodontics, University orthodontists presented with a mouth-breathing female
of the Pacific, Arthur A. Dugoni School of Dentistry, San and a CBCT minimal cross-sectional area (MCA)
Francisco, Calif.
c
Professor, Department of Orthodontics, University of the
reading of 50 mm2, 42% said they would modify their
Pacific, Arthur A. Dugoni School of Dentistry, San Francisco, treatment approach based on that information. The
Calif. most common modification chosen (69%) was to avoid
d
Professor and Chair, Department of Orthodontics, University extraction.5
of the Pacific, Arthur A. Dugoni School of Dentistry, San
Francisco, Calif.
Previous studies investigating the effect of extrac-
Corresponding author: Dr Heesoo Oh, Professor and Chair, tions on the airway have found either a reduction in
Department of Orthodontics, University of the Pacific, Arthur A. dimensions after extractions6,7 or no significant chang-
Dugoni School of Dentistry, 155 Fifth Street, San Francisco, CA es.8–10 It has been proposed11 that the indication for
94103
(e-mail: hoh@pacific.edu)
extraction may be related to how the upper airway
changes after treatment. For example, large retraction
Accepted: May 2019. Submitted: February 2019.
Published Online: August 12, 2019 of the incisors to resolve bimaxillary protrusion after
Ó 2020 by The EH Angle Education and Research Foundation, premolar extraction with maximum anchorage resulted
Inc. in decreased upper airway volume,7 whereas airway

DOI: 10.2319/021019-92.1 39 Angle Orthodontist, Vol 90, No 1, 2020

40 JOY, PARK, CHAMBERS, OH

Table 1. Reasons for Exclusion and Number of Patients Excludeda

No. of Excluded
Reasons for Exclusion Patients (n ¼ 77)
Craniofacial anomalies 3
Orthognathic surgery 32
Extraction of teeth other than premolars 7
Missing CBCT at T1 or T2 35
a
CBCT indicates cone-beam computed tomography; T1,
pretreatment; and T2, posttreatment.

dimensions increased following extractions as a result

of severe crowding with minimum anchorage space
closure.12
Previous research has produced inconclusive find-
ings. The purpose of this study was to investigate the
relationship between extraction of premolars and
changes in airway dimensions and possible associated
skeletal and dental features in a sample of nongrowing
patients. This research was undertaken to address four
questions:

1. Is extraction as part of treatment associated with Figure 1. Sampling flow chart for patient selection, with exclusion
changes in airway dimensions? criteria.
2. Is extraction as part of treatment associated with
selected skeletal and dental dimensions? extraction group and 42 subjects in the nonextraction
3. Do initial skeletal and dental dimensions interact group. Within the extraction group, six subjects had
with extraction to change airway dimensions? only upper premolars extracted, and 35 subjects had
4. Do measurement features of CBCT mask changes four premolars extracted.
in airway, skeletal, and dental features? This study was approved by the institutional review
board at the University of the Pacific (18-27).
MATERIALS AND METHODS Seventy-three CBCT scans were taken using the i-
CAT Classic unit with 40-second exposure time (23 cm
Sample selection began with identifying all patients
who were treated between January 2007 and June Table 2. Sample Characteristics and Demographic Information by
2018 in the Orthodontics Department at the University Treatment Groupa
of the Pacific, Arthur A. Dugoni School of Dentistry, Extraction Group Nonextraction Group
with appropriate CBCT images obtained as part of
Variable (n ¼ 41) (n ¼ 42)
routine diagnostic records. Subjects met the following Sex
inclusion criteria: they were at least 18 years old at Male 20 22
initiation of orthodontic treatment and there was Female 21 20
availability of pre- (T1) and posttreatment (T2) CBCTs. Age, y
,20 8 7
After exclusion criteria were applied (Table 1), 221 20–30 26 26
patients were eligible to participate in this study (Figure .30 7 9
1). To be included in the extraction group, patients had Mean age at T1, y 26.1 6 7.1 26 6 8.0
to have had at least two premolars extracted. This T1-T2, y 3.5 6 1.6 2.3 6 0.9
Treatment type
condition was determined by the fact that shortening of
Fixed appliance 40 37
the arch dimension by incisor retraction is primarily Clear aligner 1 5
achieved by premolar extraction. Subjects in the Angle classification
nonextraction group were selected by matching for Class I 16 20
age and sex with subjects in the extraction group Class II 22 20
Class III 3 2
(Table 2). This sampling process resulted in 42 Initial crowding, mm
matched subjects for the extraction and nonextraction Mild (3) 7 9
groups. One extraction subject was later excluded as a Moderate (4–6) 10 28
result of poor CBCT scan quality. Therefore, the final Severe (7) 24 5
sample of 83 subjects included 41 subjects in the a
T1 indicates pretreatment; T2, posttreatment.

Angle Orthodontist, Vol 90, No 1, 2020

AIRWAY CHANGES AFTER PREMOLAR EXTRACTIONS IN ADULT 41

Table 3. Definitions of Skeletal and Dental Landmarks by aligning the inferior border of the right and left orbits;
Landmark Symbol Definition (2) coronal plane was determined by the Frankfort
Skeletal landmarks horizontal plane (right porion and right orbitale); and (3)
Nasion N The midpoint of the frontonasal sagittal plane was adjusted by aligning the most
suture anterior point on the lateral borders of the right and
Basion Ba The most inferior and posterior left orbital rims. Skeletal and dental landmarks were
point at the anterior margin of
the foramen magnum
identified using the 3D Analysis function of Invivo
Sella S The midpoint of the cavity of software (Table 3).
sella turcica in all three planes For purposes of analysis, the airway was divided into
Orbitalea Or The most inferior point along the three regions: nasopharynx (superior to the palatal
inferior margin of the orbital rim plane), retropalatal region of the oropharynx (between
Poriona Po The most superior and lateral
point of the external auditory the palatal plane and the base of the soft palate at the
meatus most anterior point), and retroglossal region of the
Anterior nasal ANS The most anterior point of the oropharynx (between the base of the soft palate and
spine maxilla the base of the epiglottis). Airway segmentation
Posterior nasal PNS The most posterior point of the
threshold values were adjusted to eliminate imaging
spine palatine bone
Point A A The deepest point on the contour artifacts and were held constant at about 400 relative
of the maxilla between the Hounsfield units. The airway volume was then calcu-
anterior nasal spline and the lated in cubic centimeters, and the most constricted
upper incisor MCA of the airway was calculated in square millimeters
Point B B The innermost point on the
contour of the mandible
for each of the three regions using the airway function
between the lower incisor and of Invivo software. Additionally, in order to standardize
the bony chin measurements based on anatomical landmarks, the
Menton Me The most inferior point along the sagittal length (AP, in millimeters), lateral width (Lat, in
middle of the mandibular
millimeters), and cross-sectional area (XSec, in square
symphysis
Goniona Go The most inferior point of the millimeters) were measured at the level of the palatal
angle of the mandible where plane, the base of the soft palate, and the base of the
the body of the mandible epiglottis (Figure 2; Table 4). In total, 15 airway and 10
meets the ramus cephalometric measurements were determined by
Dental landmarks
U1 incisal edgea U1 The most mesial point along the
averaging the measurements from two judges (Figures
maxillary central incisor incisal 2 and 3). Table 4 summarizes the variables used in this
edge study.
U1 apexa U1A The maxillary central incisor root
apex Statistical Analysis
L1 incisal edgea L1 The most mesial point along the
mandibular central incisor All airway measurements were performed by two
incisal edge judges. The average of their values was used for
L1 apex a
L1A The mandibular central incisor
root apex analysis, and the consistency of their reading of the
U6 MB cuspa U6_MBC The maxillary first molar CBCT images was gauged by Cronbach’s alpha, a
mesiobuccal cusp tip universal ICC measure. Descriptive statistics were
L6 MB cuspa L6_MBC The mandibular first molar calculated for 15 airway and 10 skeletal and dental
mesiobuccal cusp tip
variables at T1 and T2.
a
Bilateral landmarks (right and left). Independent t-tests were performed to determine
whether groups were comparable at baseline. Repeat-
3 17 cm FOV, 0.3-mm voxel size), and 93 CBCT scans ed-measures analysis of variance (ANOVA) tests were
were taken using the i-CAT Next Generation unit with performed to test for differences across time and
8.9-second exposure time (23 cm 3 17 cm FOV, 0.3- across group (extraction vs nonextraction) and to
mm voxel size). While taking the scans, subjects were identify interactions showing that extraction differen-
sitting upright and were instructed to bite into maximal tially affected airway, skeletal, or dental features.
intercuspation and remain stationary without swallow- Repeated-measures tests were appropriate given that
ing. No instructions were given regarding the mode of failure to consider within-subject variance typically
breathing or tongue position. DICOM images were overestimates the sensitivity of tests. Correlations
imported into Invivo software (version 6; Anatomage, between baseline scores and change in score, using
San Jose, Calif) and deidentified for analysis. All scans Pearson correlation coefficients, were performed to
were oriented as follows: (1) axial plane was adjusted identify potential confounding measurement sources of

Angle Orthodontist, Vol 90, No 1, 2020

42 JOY, PARK, CHAMBERS, OH

Table 4. Airway and Skeletal and Dental Cephalometric

Measurements
Measurements Definition
Airway analysis
Anatomic defined measurements
AP, mm Sagittal linear dimension at the lowest
border of each region of the airway
Lat, mm Lateral linear dimension at the lowest
border of each region of the airway
XSec, mm2 Cross-sectional area on the axial plane at
the lowest border of each region of the
airway
Regional airway measurements
Vol, cm3 Total volume of each region of the airway
MCA, mm2 The most constricted minimal cross-
sectional area of each region of the
airway
Cephalometric analysis
Skeletal
Sperp-A, mm Horizontal distance from the
perpendicular line through Sella to
Point A
Sperp-B, mm Horizontal distance from the
perpendicular line through Sella to
Point B
FMA, 8 Mandibular plane angle; FH to mandibular
plane (Go-Me)
Figure 2. Regions of the airway and airway measurements. (A) 1, Dental
Nasopharynx; 2, Retropalatal; 3, Retroglossal; (B) (1) level of PNS; Sperp-U1, mm Horizontal distance from the
(2) level of base of soft palate; (3) level of base of epiglottis for cross perpendicular line through Sella to U1
sections parallel to Frankfort horizontal at standardized anatomical Sperp-L1, mm Horizontal distance from the
locations; (C) AP, Lat, and XSec at the level of the base of the soft perpendicular line through Sella to L1
palate. U1PPA, 8 Angle between the long axis of the U1
and the palatal plane
variance. Sequential-entry multiple regression analysis IMPA, 8 Angle between the long axis of the L1
and the mandibular plane (Go-Me)
was performed to control for potentially confounding IIA, 8 Interincisal angle; angle between the long
effects. Exact test statistics are reported, as are exact axis of the U1 and long axis of the L1
U6-6, mm Maxillary intermolar width (UR6_MBC-
P-values for all tests.
UL6_MBC)
L6-6, mm Mandibular intermolar width (LR6_MBC-
RESULTS LL6_MBC)

The average Cronbach alpha coefficient for consis-

tency in airway readings for two judges was 0.981, with
no coefficient below 0.954.

Figure 3. Three-dimensional cephalometric measurements.

Angle Orthodontist, Vol 90, No 1, 2020

AIRWAY CHANGES AFTER PREMOLAR EXTRACTIONS IN ADULT 43

Table 5. Airway Characteristics as a Function of Time and Extraction (EXT), Descriptive Means (Standard Deviations), and Tests of Hypothesis
and P-Values Concerning Patterns in the Dataa
EXT Group Non-EXT Group
(n ¼ 41) (n ¼ 42) Significance
D/Base
T1 T2 T1 T2 (Corr) Base Group Time Gr 3 T
Nasopharynx
AP, mm 19.86 19.48 21.30 20.71 .426*** 1.950þ 3.997* 1.875 0.084
(3.32) (3.57) (3.40) (3.49) ,.001 0.055 0.049 0.175 0.772
Lat, mm 26.31 26.33 26.42 26.21 .283*** 0.141 0.001 0.085 0.123
(3.56) (3.61) (3.75) (4.45) 0.01 0.888 0.995 0.77 0.727
Xsec, mm2 504.46 501.15 523.32 509.95 .406*** 0.68 0.275 0.482 0.189
(125.0) (131.3) (126.9) (133.7) ,.000 0.498 0.601 0.489 0.665
Vol, cm3 5.37 5.32 4.70 4.87 .134 1.691þ 1.968 0.259 0.851
(1.81) (2.06) (1.80) (1.92) 0.226 0.095 0.164 0.612 0.359
MCA, mm2 289.30 266.79 242.01 256.83 .475*** 2.869** 3.679þ 0.307 7.212**
(72.1) (76.0) (77.9) (73.7) ,.001 0.005 0.059 0.581 0.009
Retropalatal region of oropharynx
AP, mm 11.00 11.09 9.84 9.28 .395*** 1.66 5.521* 0.462 0.93
(3.27) (3.58) (3.08) (3.17) ,.001 0.101 0.021 0.499 0.338
Lat, mm 21.27 20.76 21.23 19.81 .605*** 0.02 0.111 1.241 0.282
(6.09) (8.18) (10.20) (6.22) ,.001 0.981 0.74 0.269 0.597
Xsec, mm2 230.18 238.27 188.98 175.80 .425*** 1.791þ 6.357* 0.044 0.77
(118.1) (134.3) (90.0) (86.0) ,.001 0.077 0.014 0.834 0.383
Vol, cm3 8.43 8.34 7.79 7.58 .299** 0.899 1.042 0.191 0.035
(3.42) (4.23) (2.99) (3.08) 0.006 0.371 0.31 0.663 0.851
MCA, mm2 182.55 167.52 149.60 139.99 .312** 1.850þ 0.038 1.724þ 0.083
(86.1) (110.9) (75.9) (83.0) 0.004 0.068 0.193 0.085 0.774
Retroglossal region of oropharynx
AP, mm 10.23 9.84 10.20 10.77 .444*** 0.108 0.561 0.068 1.951
(2.97) (3.49) (2.93) (2.74) ,.001 0.914 0.456 0.794 0.167
Lat, mm 28.14 27.84 26.29 26.50 .330** 1.461 1.467 0.005 0.145
(5.53) (7.26) (6.49) (6.55) 0.003 0.148 0.23 0.943 0.704
Xsec, mm2 241.24 230.05 209.85 216.82 .384*** 1.831þ 1.634 0.044 0.814
(79.6) (108.4) (82.7) (79.9) 0.001 0.071 0.205 0.835 0.37
Vol, cm3 6.99 7.15 6.15 6.08 .293** 1.25 1.907 0.023 0.13
(3.47) (3.96) (2.82) (2.97) 0.009 0.215 0.171 0.879 0.72
MCA, mm2 172.04 174.49 134.03 126.04 .474*** 2.007* 6.934** 0.092 0.325
(98.3) (92.7) (66.2) (70.0) ,.001 0.048 0.01 0.763 0.57
a
Airway features—AP indicates sagittal dimension measured at the inferior border of each region; Lat, transverse dimension measured at the
inferior border of each region; XSec, area measured at the inferior border of each region; Vol, volume; and MCA, minimal cross-sectional area.
Tests performed on data—D/Base indicates change in airway feature correlated with base value of that feature at pretreatment (T1) (r); Base,
difference between average feature value at T1 for extraction and nonextraction cases (t); Group, difference between average extraction and
nonextraction values combining T1 and posttreatment (T2) (F); Time, difference between values at T1 and T2 combining groups (F for repeated-
measures analysis of variance [ANOVA]); and Gr 3 T, interaction of extraction or nonextraction over time (F for repeated-measure ANOVA).
Report of tests: Test statistics (r, t, or F) reported on first line; P-value reported on second line.
þ P , .10; * P , .05; ** P , .01; *** P , .001.

The results for airway measures are shown in Table groups at baseline). In five of 15 cases, subjects in the
5. The mean and standard deviation 2 3 2 crossed extraction group showed larger airway values when
classification of scores for the five airway measures in combined across both T1 and T2 (main effect for group
the nasopharynx, retropalatal, and retroglossal regions in repeated-measures ANOVA). There was only one
are shown. Figure 4A is a graph of a representative feature (MCA in the retropalatal areas) for which that
result for retropalatal volume. This is typical of the effect showed a marginally significant difference during
findings by virtue of showing no change over time, a treatment, regardless of extraction (main effect for time
slight tendency for extraction cases to have larger in repeated-measures ANOVA). It appeared that there
measurements both initially and at the end of treat- was a differential effect of extraction in the single case
ment, and no differential effects with extraction. of MCA in the nasopharynx region. That difference
Although subjects were matched based on age and failed to reach significance at P , .10 when corrections
sex, for seven of the 15 airway features measured, the are taken for measurement error.
extraction group began the study with slightly larger The only consistent finding in these data was that
airways (independent t-tests for differences between changes in airway features were associated with their

Angle Orthodontist, Vol 90, No 1, 2020

44 JOY, PARK, CHAMBERS, OH

Figure 4. (A) Change in retropalatal volume in extraction and nonextraction groups; (B) Change in retropalatal MCA as a function of baseline
values in extraction and nonextraction cases for subjects in whom baseline MCA was less than 100 mm2 or greater than 200 mm2.

baseline conditions. For all but one feature, the cases that were larger than 200 mm2 tended to
correlation between baseline score and change score decrease in size during treatment.
was highly significant. The association was always Table 6 is a parallel summary of the analysis of three
negative, demonstrating that small airways grew in size skeletal and seven dental features as they were
and large airways decreased in size. This effect is affected by extraction or nonextraction. The positions
diagrammed in Figure 4B for retropalatal MCA, of both the maxillary and mandibular incisors were
showing that both extraction and nonextraction cases larger in the extraction group initially but decreased
that were initially smaller than 100 mm2 tended to significantly in the extraction group and increased
increase and that both extraction and nonextraction significantly in the nonextraction group. The same

Table 6. Skeletal and Dental Characteristics as a Function of Time and Extraction (EXT), Descriptive Means (Standard Deviations), and Tests of
Hypothesis and P-Values Concerning Patterns in the Dataa
EXT Group Non-EXT Group
(n ¼ 41) (n ¼ 42) Significance
D/Base
T1 T2 T1 T2 (corr) Base Group Time Group 3 T
Skeletal
Sperp-A, mm 65.81 65.79 68.81 68.73 0.04 3.223** 10.033** 0.639 0.175
(4.05) (4.09) (4.42) (4.55) 0.74 0.008 0.002 0.43 0.68
Sperp-B, mm 59.57 59.45 63.54 63.41 .114 2.816** 8.018** 0.438 0.001
(3.56) (3.61) (3.75) (4.45) 0.31 0.006 0.008 0.51 0.98
FMA, 8 25.19 25.13 22.79 22.74 0.175 1.631 2.73 0.122 0
(7.19) (7.17) (6.18) (5.97) 0.11 0.11 0.10 0.73 0.997
Dental
Sperp-U1, mm 67.64 65.39 70.46 71.31 0.11 2.367* 14.743*** 7.657** 5.047***
(5.58) (5.05) (5.26) (5.36) 0.34 0.02 ,.000 0.01 ,.000
Sperp-L1, mm 70.97 68.37 73.58 73.99 .168 2.144* 25.688*** 13.574*** 12.768***
(6.05) (5.21) (5.00) (5.38) 0.13 0.035 ,.000 ,.000 0.001
U1PPA, 8 112.48 109.74 110.57 113.45 .585*** 2.367* 0.363 0.006 8.711**
(8.98) (8.33) (7.64) (7.22) ,.001 0.02 0.55 0.94 0.004
IMPA, 8 99.39 100.30 99.69 100.77 .430*** 1.046 0.097 1.889 0.018
(6.81) (7.31) (5.20) (6.26) ,.001 0.30 0.75 0.17 0.89
IIA, 8 126.28 130.59 130.21 123.47 .712*** 1.324 0.538 0.722 14.934***
(14.16) (9.06) (12.82) (10.69) ,.001 0.19 0.47 0.40 ,.001
U6-6, mm 51.37 49.73 51.90 52.77 .491** 0.732 8.584** 2.716 28.691***
(3.39) (2.09) (3.23) (3.02) ,.001 0.47 0.004 0.10 ,.001
L6-6, mm 45.71 43.95 46.27 47.10 .368*** 0.742 7.338** 3.130þ 24.172***
(3.03) (3.07) (3.85) (3.21) 0.001 0.46 0.008 0.08 ,.001
a
Tests performed on data—D/Base indicates change in airway feature correlated with base value of that feature at pretreatment (T1) (r); Base,
difference between average feature value at T1 for extraction and nonextraction cases (t); Group, difference between average extraction and
nonextraction values combining T1 and posttreatment (T2) (F); Time, difference between values at T1 and T2 combining groups (F for repeated-
measures analysis of variance [ANOVA]); and Gr 3 T, interaction of extraction or nonextraction over time (F for repeated-measures ANOVA).
Report of tests: Test statistics (r, t, or F) reported on first line; P-value reported on second line.
þ ¼ P , 0.10; * P , .05; ** P , .01; *** P , .001.

Angle Orthodontist, Vol 90, No 1, 2020

AIRWAY CHANGES AFTER PREMOLAR EXTRACTIONS IN ADULT 45

significant interaction was observed for both the Averaging measurements across subjects masked
maxillary and mandibular intermolar widths. The conclusions about both individual and aggregate
opposite effect was seen with respect to interincisor changes in airway features. This was the largest and
angle, which indicates the impact of extraction on most consistent finding in this research and an effect
incisor inclination. Very large associations were ob- that applied equally to measures of airway and skeletal
served between baseline and change measures for and dental features. In 19 of the 25 variables analyzed,
five of the dental features representing incisor inclina- this effect was significant at P , .01. The consistent
tion and intermolar width. negative associations between baseline score and
Correlation matrices were calculated relating chang- change indicated that small values at T1 were more apt
es in the 15 airway features with both baseline and to grow in size by T2 and that initially large values
change measures for the 10 skeletal and dental consistently got smaller, which creates a chance for
features. These calculations were performed for both bias, especially if only large initial airways are studied.
the total sample and separately for only the extraction Such sampling restrictions are sometimes used on
subjects. Of the 600 coefficient values calculated, only ethical grounds to mitigate supposed risk to patients
seven of them were statistically significant. with small airways or, in general, if adequately large
airways dominate a sample.
Often, the difference between repeated measures of
DISCUSSION
a feature can be attributed to treatment changes, but it
A retrospective analysis was conducted on 83 is not always possible to isolate these effects from
nongrowing, orthodontically treated patients, matched confounding factors such as measurement artifacts
for sex and age, in which half received premolar and natural growth. In most cases, the effects of
extractions and half did not. Five features in each of measurement and growth are not random, and they
the nasopharynx, retropalatal, and retroglossal airway interact with planned interventions.
regions and 10 skeletal and dental features were Measurement error across repeated measures is
assessed at the beginning and end of treatment. known as ‘‘regression toward the mean.’’15 Values far
Results were analyzed across time, across treatment from the mean at T1 are seen to be closer to the mean
type, and for interactions. Associations were also at T2 because of the underlying nature of random
explored between airway and skeletal and dental sampling. It is unlikely that extreme values will become
measures. even more extreme. The more unreliable the measure-
The results supported the conclusion that extraction ment system, the larger the regression toward the mean.
has an effect on dental features such as upper and lower This effect can be approximated by looking for negative
incisor position and inclinations and intermolar width. correlations between baseline score and change from
However, there was no evidence that extraction changed T1 to T2 or by calculating the ICC with the difference
between individual scores at baseline and baseline
sagittal and transverse distances or minimal cross-
average as a covariate. This study showed an extremely
sectional area or volume in the nasopharyngeal, retro-
high consistency in the reading of CBCT images of
palatal, or retroglossal regions. In addition, there was no
airway features but significant and consistent inconsis-
evidence that changes in the measured skeletal or dental
tency in the taking of these images. Postural and
features had an indirect effect on airway features. Strong
functional fluctuations from one exposure to the next can
evidence emerged that common measurement effects in
vary depending on the type of feature being measured.
repeated measures may obscure the types of conclu- The conclusions of this study can only be general-
sions drawn from similar clinical cases. ized to nongrowing patients and are limited to
Two early studies6,7 that reported a negative impact conclusions measured by the variables chosen. A
of extraction on airway did not use CBCT technology, large number of zero-order correlations were calculat-
so they could not report transverse dimensions, area, ed between skeletal and dental features and charac-
or volume. They reported correlations on changes teristics of the airway, and they were found to be
rather than differences across groups and did not use a insignificant. However, it remains possible that patterns
control group. Pliska et al.10 reported results similar to of several skeletal and dental features might have an
those found in the current study. Their sample effect. Multiple regression and latent structure analy-
contained patients with larger initial airways and very ses may prove useful in clarifying this relationship. It
large standard deviations,10 but their results were was clear that factors such as planned therapeutic
similar in finding regression toward mean values. With effects, natural growth, and the interaction of anatom-
respect to skeletal and dental features, the findings in ical features are confounded with the variable timing of
the current study were similar to those reported growth across individuals and the statistical complica-
previously in the literature.9,12–14 tions of using measurement systems with significant

Angle Orthodontist, Vol 90, No 1, 2020

46 JOY, PARK, CHAMBERS, OH

potential for unreliability across repeated measures. 4. Keim RG, Gottlieb EL, Vogels DS 3rd, Vogels PB. 2014 JCO
Much work remains to be done to disentangle these Study of Orthodontic Diagnosis and Treatment Procedures:
sources of variation in order to prevent mistaking the part 1: results and trends. J Clin Orthod. 2014;48:607–630.
5. Cummings S. Diagnostic Paths for a Mouth-Breathing
reasons for observed changes or, in this case, the
Patient [unpublished MSD thesis]. San Francisco, Calif:
absence of changes, averaging across patients. University of the Pacific, Arthur A. Dugoni School of
The assessment of airway dimensions utilizing Dentistry; 2018.
CBCTs is subject to many limitations. For example, 6. Wang Q, Jia P, Anderson NK, Wang L, Lin J. Changes of
body mass index was not recorded, which might affect pharyngeal airway size and hyoid bone position following
the airway dimensions. In addition, patient positioning orthodontic treatment of Class I bimaxillary protrusion. Angle
and the process of breathing have been shown16–19 to Orthod. 2012;82:115–121.
change upper airway volume, size, and shape. 7. Chen Y, Hong L, Wang CL, et al. Effect of large incisor
Patients with narrow upper airway cross sections can retraction on upper airway morphology in adult bimaxillary
protrusion patients. Angle Orthod. 2012;82:964–970.
maintain airway patency by dilating their airway,20
8. Valiathan M, El H, Hans MG, Palomo MJ. Effects of
illustrating that the dimensions of the airway are extraction versus non-extraction treatment on oropharyngeal
dynamic and variable. airway volume. Angle Orthod. 2010;80:1068–1074.
9. Al Maaitah E, El Said N, Abu Alhaija ES. First premolar
CONCLUSIONS extraction effects on upper airway dimension in bimaxillary
proclination patients. Angle Orthod. 2012;82:853–859.
 Although the indications for extractions in orthodon-
10. Pliska BT, Tam IT, Lowe AA, Madson AM, Almeida FR.
tics have been debated, the fact remains that some Effect of orthodontic treatment on the upper airway volume
cases necessitate removal of teeth in order to in adults. Am J Orthod Dentofacial Orthop. 2016;150:937–
achieve treatment and/or esthetic goals. As a result 944.
of the variability in airway dimensions, it is advised to 11. Hu Z, Yin X, Liao J, Zhou C, Yang Z, Zou S. The effect of
consider additional diagnostic information, such as teeth extraction for orthodontic treatment on the upper airway:
clinical examination, sleep questionnaires, and poly- a systematic review. Sleep Breath. 2015;19:441–451.
somnography, rather than relying solely on airway 12. Germec-Cakan D, Taner T, Akan S. Uvulo-glossopharyn-
measurements from a CBCT scan, when making geal dimensions in non- extraction, extraction with minimum
anchorage, and extraction with maximum anchorage. Eur J
treatment decisions regarding airway concerns.
Orthod. 2011;33:515–520.
 Claims that extraction treatment is generally contra- 13. Zhang J, Chen G, Li W, Xu T, Gao X. Upper airway changes
indicated because of airway considerations are too after orthodontic extraction treatment in adults: a preliminary
general to be useful. Extraction cases should be study using cone beam computed tomography. PLoS ONE.
evaluated on a case-by-case basis, preferably with 2015;10:1–14.
the aid of additional diagnostic evaluations, such as 14. Herzog C, Konstantonis D, Konstantoni N, Eliades T. Arch-
an obstructive sleep apnea screening. Characteris- width changes in extraction vs nonextraction treatments in
tics of the airway at baseline must be taken into matched Class I borderline malocclusions. Am J Orthod
account, as must the possibility of wide variation due Dentofacial Orthop. 2017;151:735–743.
15. Barnett AG, van der Pols JC, Dobson AJ. Regression to the
to the unreliability of CBCT imaging in evaluating
mean: what it is and how to deal with it. Int J Epidemiol.
certain features. Due partly to patient selection, 2005;34:215–220.
natural growth, and measurement effects, small 16. Pevernagie DA, Shepard JW Jr. Effects of body position on
airway dimensions at baseline will appear to upper airway size and shape in patients with obstructive
increase, and large airway dimensions will appear sleep apnea. Acta Psychiatr Belg. 1994;94:101–103.
to decrease after treatment. 17. Yildirim N, Fitzpatrick MF, Whyte KF, Jalleh R, Wightman
AJ, Douglas NJ. The effect of posture on upper airway
dimensions in normal subjects and in patients with the sleep
REFERENCES apnea/hypopnea syndrome. Am Rev Respir Dis. 1991;144:
1. Avrahami E, Englender M. Relation between CT axial cross- 845–847.
sectional area of the oropharynx and obstructive sleep 18. El H, Palomo JM. Measuring the airway in 3 dimensions: a
apnea syndrome in adults. Am J Neuroradiol. 1995;16:135– reliability and accuracy study. Am J Orthod Dentofacial
140. Orthop. 2010;137(suppl):S50.e1–S50.e9.
2. Ogawa T, Enciso R, Shintaku WH, Clark GT. Evaluation of 19. Tso HH, Lee JS, Huang JC, Maki K, Hatcher D, Miller AJ.
cross-section airway configuration of obstructive sleep Evaluation of the human airway using cone-beam comput-
apnea. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. erized tomography. Oral Surg Oral Med Oral Pathol Oral
2007;103:102–108. Radiol Endod. 2009;108:768–776.
3. Vos W, De Backer J, Devolder A, et al. Correlation between 20. Cheng S, Brown EC, Hatt A, Butler JE, Gandevia SC, Bilston
severity of sleep apnea and upper airway morphology based L. Healthy humans with a narrow upper airway maintain
on advanced anatomical and functional imaging. J Biomech. patency during quiet breathing by dilating the airway during
2007;40:2207–2213. inspiration. J Physiol. 2014;592:4763–4774.

Angle Orthodontist, Vol 90, No 1, 2020