ORIGINAL ARTICLE

The effects of maxillary protraction therapy

with or without rapid palatal expansion:
A prospective, randomized clinical trial
Gregory A. Vaughn,a Brian Mason,b Hong-Beom Moon,c and Patrick K. Turleyd
Seattle, Wash, Phoenix, Ariz, and Los Angeles, Calif

Introduction: The purpose of this controlled randomized clinical trial was to quantify the effects of maxillary
protraction with or without palatal expansion. Methods: Forty-six children aged 5 to 10 years were randomly
assigned to 1 of 3 groups: (1) facemask with palatal expansion, (2) facemask without palatal expansion, and
(3) observation for 12 months. Cephalometric analysis with traditional cephalometric measurements, an x-y
coordinate system, and an occlusal-plane analysis were used. Results: Student t tests showed no significant
differences (P ⬍ .05) between expansion and nonexpansion groups in any measured variable. Comparisons
of treated and control subjects showed significant (P ⬍ .01) treatment effects beyond normal Class III growth.
Analysis of x-y coordinate variables showed the following: 2 mm additional forward displacement of the
maxillary complex with counterclockwise rotation, mandibular clockwise rotation, posterior movement of
B-point by an average of 1.5 mm, and forward movement of the maxillary dentition of nearly 1 mm. Analysis
of traditional cephalometric measures showed improvements in ANB angle of nearly 4° and Wits appraisal
of nearly 4 mm. The occlusal plane analysis showed an apical base change of 4 mm, 1.5 mm forward
displacement of the maxillary complex, mandibular clockwise rotation of 2.5 mm, and forward maxillary
molar movement of 1.9 mm. Conclusions: The results of this continuing 5-year clinical trial indicate that early
facemask therapy, with or without palatal expansion, is effective to correct skeletal Class III malocclusions.
(Am J Orthod Dentofacial Orthop 2005;128:299-309)

E
arly management of Class III malocclusions
with maxillary protraction has become common
in the orthodontic community. A developing
Class III malocclusion can include maxillary skeletal
retrusion, mandibular skeletal protrusion, or a combi-
nation of both.1 A posterior or anterior crossbite is also
frequent.2,3 Common dental compensations might in-
clude maxillary dentoalveolar protrusion and mandib-
ular dentoalveolar retrusion.4,5 In addition to a promi-
nent chin, a midface deficiency is often apparent, as
well as increased mandibularr incisor display and a
prominent lower lip.6
a
Private practice, affiliate assistant professor, Department of Orthodontics,
School of Dentistry, University of Washingtion, Seattle, Wash.
b
Private practice, Phoenix, Ariz.
c
Associate clinical professor, Section of Orthodontics, School of Dentistry,
University of California at Los Angeles.
d
Adjunct professor emeritus, Section of Orthodontics, School of Dentistry,
University of California at Los Angeles.
In partial fulfillment of the requirements for Certificate in Orthodontics at the
University of California at Los Angeles (first author).
Supported by the American Association of Orthodontists Foundation and the
Foundation for Orthodontic Research.
Reprint requests to: Dr Gregory A. Vaughn, 215 1st Ave W, Suite 100, Seattle,
WA 98119; email, Greg@LandVortho.com.
Submitted, August 2003; revised and accepted, April 2005.
0889-5406/$30.00
Copyright © 2005 by the American Association of Orthodontists.
doi:10.1016/j.ajodo.2005.04.030
Historically, the Class III skeletal malocclusion was
viewed as a problem of the mandible. Until the 1970s,
Class III and mandibular prognathism were virtually
synonymous.7 Many studies since have found that, in
most cases, the maxilla is also hypoplastic and often is
the primary etiology of the Class III malocclusion.
Clinical studies focusing on maxillary protraction1,8-26
report anterior displacement and counterclockwise ro-
tation of the maxilla, downward and backward rotation
of the mandible, forward movement of the maxillary
dentition, retroclination of the mandibular incisors, and
increased lower face height.
A recent study evaluating skeletal maturity as an
indicator for the timing of maxillary protraction showed
that treatment effects could be more pronounced in
younger patients.24 Several recent studies, evaluating
the soft tissue changes due to maxillary protraction,
showed increases in convexity and improved upper lip
positions leading to improvements in the profile.19,27
These studies are difficult to compare because of
differences in appliance design, use of palatal expan-
sion in some subjects, location of protraction force
application, subjects’ ages, duration of treatment, treat-
ment end points, and controls. In addition, methods of
data analysis are not standardized, and cephalometric
superimposition and analysis vary across the studies.
299
300 Vaughn et al
Table I. Summary statistics for mean ages, sex, and average treatment time
Group n Mean age
A (expansion) 15 7.3833
B (nonexpansion) 14 8.1086
C (observation) 17 6.6265
SE, Standard error.
Control groups are absent in many studies, and, when
present, they are mainly historical Class I or III con-
trols. Ngan et al12,15 and Merwin et al20 performed
prospective trials, with each subject serving as his or
her own control by using a pretreatment 6-month period
for observing growth, but no randomization was used in
the design.
Maxillary expansion has been advocated as an
important part of maxillary protraction with facemask
therapy.1,6 The effects of rapid palatal expansion have
been well documented in animal and clinical trials.
Rapid palatal expansion is easily accomplished in a
growing patient and leads to downward and forward
movement of A-point of approximately 1.5 mm, more
significant widening at the canine-premolar region than
at the molar region, and downward and backward
rotation of the mandible, thus increasing lower face
height.2,28-30 Some perceived benefits are associated
with rapid maxillary expansion in conjunction with
maxillary protraction therapy in treating skeletal Class
III patients, including transverse expansion to correct
crossbites often associated with Class III malocclu-
sions, splinting of the maxillary dentition against for-
ward movement and anterior constriction during pro-
traction therapy, backward and downward rotation of
the mandible, and disarticulation of the circummaxil-
lary sutures. This disarticulation is postulated to prime
the sutures for more pronounced orthopedic ef-
fects.6,15,17,23,31
To date, there is no report in the literature of a
prospective, controlled, randomized clinical trial on
Class III treatment with maxillary protraction. There-
fore, the purpose of this study was to quantify the
effects of maxillary protraction by using a custom-
ized facemask with or without rapid palatal expan-
sion in the early correction of the skeletal Class III
malocclusion.
MATERIAL AND METHODS
The patient sample was recruited by advertising to
community dentists, in schools and local newspapers,
and on the radio. No effort was made to select patients
based on sex or ethnic background. The patients’ ages
ranged from 5 to 10 years. Age, sex, and treatment
American Journal of Orthodontics and Dentofacial Orthopedics
September 2005
SE Female Male Average time (y)
0.5035 8 7 1.16
0.5212 7 7 1.15
0.473 7 10
times are summarized in Table I. The total sample size
was 46 patients.
The inclusion criteria were zero or negative overjet
on 2 or more incisors, Class III molar relationship with
the mesiobuccal cusp of the maxillary permanent first
molar distal to the buccal groove of the mandibular first
permanent molar, or a mesial step terminal plane
relationship of 3.0 mm or more if the deciduous molars
were present (measured clinically). When the clinical
or dental criteria were borderline, we used cephalomet-
ric criteria of ANB angle of 0° or less, Wits analysis of
3 mm or more, and nasion perpendicular to A-point of
2 mm or less. Exclusion criteria were any craniofacial
anomaly, psychosocial impairment, or skeletal open
bite.
We used a block randomization table to assign the
subjects to 1 of 3 groups after obtaining proper in-
formed consent. The principal investigator (G.A.V.)
was blinded to the assignment.
Patients in group A (n ⫽ 15 ⫹ 6) were treated with
palatal expansion with facemask therapy. Patients in
group B (n ⫽ 14 ⫹ 8) were treated with passive palatal
appliances with facemask therapy. Patients in group C
(n ⫽ 17) were observed only for at least 12 months and
then randomly assigned to group A (n ⫽ 6) or B (n ⫽
8). (Three patients who completed the observation
period were still in treatment when the records for this
study were collected.)
A banded, soldered, jackscrew palatal expansion
appliance was used for each subject in group A. Two
teeth per side were banded: the first and second
deciduous molars, the first permanent molar and the
second deciduous molar, or the first permanent molar
and premolar. The appliance was activated twice daily
(0.5 mm/day) for a minimum of 7 days. Soldered hooks
(.045 in) were extended to the mesial of the canine for
attachment of the force-delivering elastics.
The protocol in group B was identical to that for
group A except that the palatal expander was not
activated. If patients required transverse expansion, this
was performed after final records (T2) were obtained.
Group C was the control or observation group.
Initial records (T0) were taken at enrollment and 1 year
later (T1), at assignment to group A or B.
American Journal of Orthodontics and Dentofacial Orthopedics
Volume 128, Number 3
Each facemask was fabricated on a model made
from an impression of the patient’s face, according to
Turley.6 The facemask was fitted 7 to 10 days after
the placement of the palatal appliance. Elastics,
directed 15° to 30° downward from the occlusal
plane, delivered a force of 300 to 500 g per side, as
determined by a force gauge. The subjects were
instructed to wear the appliance full time at the
beginning of treatment. Compliance was closely
monitored with timecards.
Once positive overjet and overbite and Class I
molar occlusion were obtained, facemask wear was
reduced to 14 hours a day. In anticipation of some
relapse, overcorrection approaching an end-to-end mo-
lar relationship and overjet of 4 to 5 mm were the
treatment objectives. The treatment results were main-
tained for 3 to 6 months with nighttime wear. All
clinical treatment was performed by Drs Patrick K.
Turley, Hong-Beom Moon, and their designees.
Lateral cephalometric radiographs were taken at
T0, T1, and T2 for the control group, and at T1 and T2
for the 2 treatment groups. The same cephalometer was
used for all films to ensure standardization. All films
were taken with relaxed lips and in centric occlusion.
The head was naturally positioned as determined by
using a gravity-held vertical chain and having the
patients look into their own eyes in a mirror during the
imaging procedure.
All cephalograms were traced and digitized by 1
investigator (G.A.V.) under optimal conditions with a
0.5-mm lead pencil on dimensionally stable cephalo-
metric tracing acetate. Radiographs were traced side by
side to maintain consistent identification of landmarks.
The midpoint between the right and left traced images
was used for the bilateral landmarks. Fifty-five standard
cephalometric landmarks were digitized in a predeter-
mined order with a digitizer accurate to 0.001 mm. Five
sets of radiographs (5 T1 and 5 T2 cephalograms) were
chosen at random, retraced, superimposed, digitized in
duplicate, and subjected to Dahlberg’s formula,32 the
Pearson correlation, and the Spearman correlation to
establish the level of tracing and digitizing error (reli-
ability). The overall Dahlberg standard deviation was
0.61, with an average error of 0.3. The Pearson and
Spearman correlations were all above 95% (P ⬍ .001).
Traditional cephalometric measurements were used
to describe changes between pretreatment, posttreat-
ment, and control lateral cephalograms. Measurements
included a combination of the Steiner, McNamara,
Ricketts, Riedel, and Wits analyses. Changes in 55
landmarks were also evaluated relative to an x-y
coordinate system. The horizontal axis was the sella-
nasion line rotated downward 7°, and the vertical axis
Vaughn et al 301
was constructed by passing a line perpendicular to the
horizontal axis through sella. Changes were measured
according to differences in landmark position from
either T0 to T1 or T1 to T2. The Johnston analysis33
also was used to differentiate between skeletal and
dental changes and to provide a method to evaluate the
combined treatment effects (skeletal and dental) along
the mean functional occlusal plane. This analysis mea-
sures movement of the molars relative to basal bone
and movement of the jaws with respect to both the
cranial base and one another. Cephalometric tracings
were superimposed on the maxilla by orienting on the
palatal plane (ANS-PNS), the lingual palatal curvature,
and the lower anterior border of the zygomatic process
(key ridge), registering at the inferior border.33 All
sagittal measurements were measured along the mean
functional occlusal plane (average of T0 and T1, or T1
and T2, functional occlusal planes). Anteroposterior
change in the maxilla was calculated by measuring the
difference in position of spheno-ethmoidol registration
point (intersection of the greater wing of the sphenoid
with the superior border of the cranial base) with
regional superimposition on the maxilla. Mandibular
change was the change measured at the mandibular
symphysis (constructed D point) relative to the cranial
base. The net change of the maxilla to the mandible was
the apical base change (ABCH). Total molar correction
(6 of 6) was the sum of the ABCH and respective
maxillary and mandibular tooth movements. Signs of
the values were defined so that positive signs contrib-
uted to correction of a Class III relationship and
negative signs detracted from the correction. The pitch-
fork diagram was used to show differences in the
occlusal-plane analysis between the groups (Fig 1).
The subjects were required to keep a standardized
record of facemask wear throughout treatment. This
timecard was checked at each appointment to verify
compliance. Facemask wear was calculated for the first
6 months of active treatment for each subject. Statisti-
cal analysis showed a positive correlation between
amount of facemask wear and treatment results. The
statistical significance was not analyzed for this article.
Data sets were annualized to determine annual rate
differences between the groups. The arithmetic median,
mean, standard deviation, and standard error were
calculated for each variable.
Differences between the groups were determined as
appropriate with t tests and analysis of variance. The
significance level was determined at P ⬍ .05. The sexes
were combined because statistical significance was not
demonstrated between them.
302 Vaughn et al
Fig 1. Pitchfork analysis, described by Johnston,49
shows anteroposterior skeletal and dental components
of maxillary and mandibular molars. Algebraic sum of
various skeletal and dental changes equals treatment
change in molar relationship.
RESULTS
There were no statistically significant differences
between groups A and B in any measured cephalomet-
ric variable. These groups, however, had significant
skeletal changes (P ⬍ .05) relative to group C. After
subtracting the Class III growth obtained from our
observation group, we found the following in groups A
and B, respectively: the maxilla (Table II) moved
forward (SNA angle, ⫹3.02° and ⫹2.78°; maxillary
depth, ⫹2.62° and ⫹2.66°; nasion perpendicular to
A-point, ⫹2.41 and ⫹2.49 mm; A-point (horizontal),
⫹2.63 and ⫹2.48) in combination with rotation of the
palatal plane (SN-palatal plane, ⫺1.27° and ⫺1.44°).
The mandible (Table III) was displaced backward
(SNB angle, ⫺0.86° and ⫺1.23°; nasion perpendicular
to pogonion, ⫺2.15 and ⫺2.59 mm; B-point [horizon-
tal], ⫺1.41 and ⫺2.15 mm). The combined movement
of the maxilla and the mandible (Table IV) produced
ANB angle changes of ⫹3.87° and ⫹3.99° and Wits
changes of ⫹3.89 and ⫹3.74 mm (P ⬍ .001). In
addition, significant (P ⬍ .05) increases in the vertical
measurement (Go-Gn-Sn, ⫹0.91° and ⫹1.15°) were
observed.
There were no statistically significant differences
between the treatment groups in any measured variable
along the mean functional occlusal plane (Johnston
analysis) (Table V, Figs 2-4). However, groups A and
B demonstrated significant skeletal changes (P ⬍ .01)
American Journal of Orthodontics and Dentofacial Orthopedics
September 2005
relative to group C. After subtracting the Class III
growth obtained from group C, we found the following:
the ABCH (total skeletal correction) measurements
were ⫹4.01 and ⫹4.41 mm for groups A and B,
respectively. The maxilla moved forward relative to the
cranial base (⫹1.67 and ⫹1.41 mm), accounting for
about a third of the total skeletal correction. The
mandible rotated posteriorly relative to the cranial base
(⫹2.34 and ⫹3.01 mm), accounting for about two
thirds of the skeletal change. The maxillary first molar
moved forward, also assisting the Class III correction
(⫹1.62 and ⫹2.34 mm, respectively). The mandibular
dentition (measured at the first molars) moved forward,
possibly due to loss of E space, hindering our Class III
correction (⫺1.84 and ⫺2.37 mm). Total Class III
correction, according to Johnston’s pitchfork analysis,
was about 4 mm (⫹3.69 mm in group A and ⫹4.35 mm
in group B).
DISCUSSION
Palatal expansion has been advocated as a routine
part of Class III correction with facemask therapy.1,6
The benefits of palatal expansion have been claimed to
be expansion of a narrow maxilla and correction of a
posterior crossbite, increase in arch length, bite open-
ing, loosening or activation of the circumaxillary su-
tures, and initiation of downward and forward move-
ment of the maxillary complex.6 Oppenheim34 was one
of the first to observe this phenomenon, and Haas3 later
showed that maxillary expansion always moves the
maxilla downward and often moves it forward. Clini-
cians have advocated maxillary expansion a week
before starting facemask use, even in the absence of
maxillary constriction or crowding. Critical evaluation
of expansion in conjunction with maxillary protraction
has been very limited. Baik8,35 analyzed facemask
therapy using a passive, rigid labiolingual appliance
and an active rapid palatal expander. It is difficult to
compare Baik’s results with ours because
(1) in Baik’s study, different appliances were used for
the expansion and nonexpansion groups;
(2) a different treatment end point was used (Baik’s
treatment objective was negative overjet correc-
tion; ours was overcorrection to 4-5 mm of posi-
tive overjet);
(3) Baik’s subjects were older than ours;
(4) no randomization was done or information about
the validity of comparing the 2 groups was given
in his study; and
(5) there were differences in requested facemask wear.
Baik requested 12 hours per day, but our subjects
were required to wear the facemask full time until the
American Journal of Orthodontics and Dentofacial Orthopedics Vaughn et al 303
Volume 128, Number 3

Tables II. Traditional maxillary cephalometric and x-y coordinate results

Group C Group A Group B AvB CvA CvB

Mean Mean Mean Mean diff SE Mean diff SE Mean diff SE

Maxilla anteroposterior
SNA ⫺0.24 2.77 2.51 0.24 NS 0.65 3.02** 0.68 2.78** 0.67
Maxillary depth 0.32 2.95 2.98 0.03 NS 0.61 2.62** 0.65 2.66** 0.64
ANS 0.7 3.24 2.88 0.36 NS 0.56 2.54** 0.59 2.18** 0.59
A-point 0.66 3.29 3.09 0.20 NS 0.55 2.63** 0.58 2.43** 0.58
Nasion perpendicular to 0.33 2.74 2.82 0.07 NS 0.55 2.41** 0.59 2.49** 0.58
A-point
Maxillary length 1.19 4.29 3.93 0.37 NS 0.74 3.11** 0.79 2.73** 0.79
Maxilla vertical
PNS ⫺1.15 ⫺.11 ⫺1.7 0.40 NS 0.35 ⫺0.96* 0.38 ⫺0.55 NS 0.38
ANS ⫺1.6 ⫺1.32 ⫺1.02 0.30 NS 0.45 0.28 NS 0.47 0.58 NS 0.47
A-point ⫺0.52 ⫺0.99 ⫺1.28 0.29 NS 0.64 ⫺0.47 NS 0.68 ⫺0.77 NS 0.68
SN-palatal plane 0.54 ⫺0.72 ⫺0.89 0.18 NS 0.53 ⫺1.27* 0.56 ⫺1.44* 0.56

*P ⬍ .05; **P ⬍ .01; NS, not significant; diff, difference; SE, standard error.

Tables III. Traditional mandibular cephalometric and x-y coordinate results

C A B AvB CvA CvB

Mean Mean Mean Mean diff SE Mean diff SE Mean diff SE

Mandible anteroposterior
SNB ⫺0.2 ⫺1.06 ⫺1.43 0.24 NS 0.34 ⫺0.86* 0.37 ⫺1.23** 0.37
Facial depth 0.58 ⫺0.63 ⫺0.89 0.26 NS 0.37 ⫺1.21** 0.39 ⫺1.47** 0.39
Nasion perpendicular to pogonion 0.99 ⫺1.18 ⫺1.62 0.43 NS 0.62 ⫺2.15** 0.67 ⫺2.59** 0.66
Mandibular length 2.81 2.51 1.71 0.80 NS 0.54 ⫺0.30 NS 0.57 ⫺1.10 NS 0.57
B-point 0.38 ⫺1.03 ⫺1.78 0.74 NS 0.57 ⫺1.41* 0.6 ⫺2.15** 0.6
Pogonion 0.75 ⫺0.85 ⫺1.87 1.01 NS 0.61 ⫺1.61* 0.65 ⫺2.62** 0.65
Mandible vertical
GO-GN-SN 0.2 1.12 1.35 0.24 NS 0.39 0.91* 0.42 1.15* 0.41
Lower facial height ⫺0.02 0.272 1.24 0.97 NS 0.72 0.29 NS 0.77 1.26 NS 0.76
B-point ⫺2.83 ⫺3.23 -2.4 0.84 NS 1 ⫺0.24 NS 1.04 0.61 NS 1.03
Pogonion ⫺2.82 ⫺2.88 ⫺3.1 0.22 NS 0.73 ⫺0.06 NS 0.77 ⫺0.28 NS 0.77

*P ⬍ .05; **P ⬍.01; NS, not significant; diff, difference; SE, standard error.

Table IV. Traditional maxillary-mandibular cephalometric and x-y coordinate results

C A B AvB CvA CvB

Mean Mean Mean Mean diff SE Mean diff SE Mean diff SE

Maxilla/mandible anterior/posterior
ANB angle ⫺0.05 3.82 3.95 0.13 NS 0.65 3.87*** 0.7 3.99*** 0.69
Mx/Mn ifference 1.62 ⫺1.79 ⫺2.22 0.43 NS 0.76 ⫺3.40** 0.81 ⫺3.84** 0.8
Wits ⫺0.06 ⫺3.98 ⫺3.82 0.16 NS 0.75 ⫺3.89
Convexity ⫺0.16 3.35 3.7 0.35 NS 0.58 3.51*** 0.62 3.86*** 0.62

*P ⬍ .05;**P ⬍.01; ***P ⬍.001; NS, not significant; diff, difference; SE, standard error.
304 Vaughn et al American Journal of Orthodontics and Dentofacial Orthopedics
September 2005

Table V. Johnston occlusal plane analysis

C A B AvB CvA CvB

Mean Mean Mean Mean diff SE Mean diff SE Mean diff SE

ABCH ⫺0.96 3.05 3.46 0.41 NS 0.6 4.01** 0.65 4.41** 0.64
Mx to CB 1.18 2.85 2.58 0.54 NS 0.35 1.67** 0.33 1.41** 0.29
Mn to CB ⫺2.14 0.2 0.87 0.87 NS 0.78 2.34* 0.78 3.01** 0.74
Mx 6 value 1.09 2.71 3.43 0.72 NS 0.49 1.62* 0.53 2.34** 0.52
Mn6 value 0.62 ⫺1.22 ⫺1.75 0.53 NS 0.49 -1.84* 0.53 ⫺2.37** 0.52
Overall 6 value 0.84 4.53 5.19 0.66 NS 0.6 3.69** 0.64 4.35** 0.64

*P ⬍ .0; **P ⬍.001; NS, not significant; diff, difference; SE, standard error.

Fig 2. Pitchfork diagram showing mean differences between groups A and B along functional
occlusal plane (Johnston analysis).

treatment objectives were met and then 14 hours per
day for the next 3 to 6 months. Baik reported
statistically significant differences between the 2
groups with greater forward movement (1.0 mm) and
downward movement (0.5 mm) for the expansion
group at A-point. This difference between his treat-
ment groups could be attributed to the differences in
appliance designs between the 2 groups.
Our results showed no statistically significant dif-
ferences between groups A and B in any measured
variable. We also found no statistically significant
differences in overall treatment time or in the time it
took to achieve anterior crossbite correction. Our re-
sults suggest that, in the absence of objective reasons
for expansion such as maxillary width or space defi-
ciencies, expansion does not aid the correction of Class
III malocclusions with facemask therapy.
Control groups allowed us to differentiate treatment
effects from normal growth. A consideration in evalu-
ating the treatment effects of protraction facemasks by
comparison with a control group is that one expects a
Class III patient to have different maxillary or mandib-
ular growth than a Class I normal subject,26 with the
Class III patient showing less forward movement of
A-point and greater forward movement of the mandi-
ble. Macdonald et al26 compared posttreatement (face-
mask) growth with 2 control groups, a Class I and a
Class III group, retrospectively. They recommended
overcorrecting Class III patients because of different
maxillary growth in the Class III groups. By comparing
the annualized mean changes of our Class III growth
sample with those of Macdonald et al26 (Class I and
Class III subjects) and Kapust et al10 (Class I control
subjects), we also found differences in the way Class I
and Class III children tend to grow (Table VI). A-point
and ANS moved more anteriorly in the Class I group;
B-point moved more anteriorly in the Class III group.
The Wits analysis also became more Class III in the
Class III observation group.
Ngan et al12-15 used treated subjects as their own
American Journal of Orthodontics and Dentofacial Orthopedics Vaughn et al 305
Volume 128, Number 3

Fig 3. Pitchfork diagram showing mean differences between groups A and C along functional
occlusal plane (Johnston analysis).

Fig 4. Pitchfork diagram showing mean differences between groups B and C along functional
occlusal plane (Johnston analysis).

growth control group. This was done by measuring
patients’ growth for a period of time (usually 6 months)
before starting facemask therapy. Because growth in
children is a nonlinear function of age, there are
limitations to relying on data from treated patients as
their own growth control group. In our study, to ensure
an adequate sample size for statistical analysis, a
percentage (32%) of the treated subjects (groups A and
B) were initially in group C for 12 months before
random assignment to either treatment group.
Many previous studies on facemask therapy have
either had no control group,20,36 a retrospective Class I
group,10,26 a retrospective Class III group.23,26 or, at
best, a nonrandomized short-term observation period
before beginning the clinical trial.12,15,20 This study
included a group of prospective, randomized Class III
306 Vaughn et al American Journal of Orthodontics and Dentofacial Orthopedics
September 2005

Table VI. Annualized mean changes for Class I and Class III control groups
Author SNA SNB ANB Wits Mx Convexity A-point B-point ANS

Class I control groups, annualized mean changes

Macdonald26 ⫺0.3 ⫺0.36 0.6 1.05 0.48 0.88 0.48 1.26
Kapust10 ⫺0.09 0.12 ⫺0.21 ⫺0.07 ⫺0.21 1.07 n/a 1.07
Class III control groups, annualized mean changes
Vaughn ⫺0.24 ⫺0.2 ⫺0.05 ⫺0.06 ⫺0.16 0.66 0.38 0.7
Macdonald26 0.02 0.54 ⫺0.52 ⫺0.66 ⫺0.66 0.41 1.23 0.71

Fig 5. Occlusal plane diagram demonstrating formation of mean functional occlusal plane (MFOP),
and effects on position of maxilla (registered at SE perpendicular) at T1 and T2.

patients as the observation group. The availability of
this group allowed us to differentiate treatment effects
from normal Class III growth in a controlled environ-
ment.
Several studies,9,10,26 including this one, have ana-
lyzed maxillary protraction by using several measure-
ment systems to enhance cross-study comparisons.
There has been a consistent difference in the amount of
maxillary protraction reported when comparing results
from traditional cephalometrics, an x-y coordinate sys-
tem, and the Johnston occlusal-plane analysis.33 This
analysis, which measures dental and skeletal movement
along the mean functional occlusal plane, consistently
shows about 1 mm less maxillary movement than
traditional cephalometric meaures. We demonstrated in
this study, as previously reported,11,14,15,23,26,32,37,38
that the palatal plane rotates significantly in a counter-
clockwise direction with facemask therapy. Because
the occlusal plane analysis uses a mean functional
occlusal plane (average of the T1 and T2 occlusal
planes), the counterclockwise rotation recorded during
treatment places the mean functional occlusal plane in
a counterclockwise position relative to T1 and in a
clockwise position relative to T2. The change in the
position of the MFOP, relative to the occlusal plane,
moves the maxillary registration point (perpendicular
to SE point) for T1 more anteriorly. Conversely, the
maxillary registration point at T2 is recorded more
posteriorly. These compensations tend to place the
maxilla, relative to SE perpendicular, in a more anterior
position at T1, and then more posteriorly at T2,
underestimating its position, resulting in a smaller
measured anteroposterior change in the maxilla’s posi-
tion (Fig 5).
Our study demonstrated a significant response to
facemask therapy, with or without palatal expansion.
Skeletal change was a combination of anterior and
vertical movement of the maxilla, with corresponding
downward and backward rotation of the mandible.
When the annualized change was compared with the
Class III control group and the control growth sub-
tracted, the results confirmed a true skeletal effect
resulting from treatment (SNA angle, ⫹3.02° and
⫹2.78°; nasion to A-point, ⫹2.41 and ⫹2.49 mm;
American Journal of Orthodontics and Dentofacial Orthopedics
Volume 128, Number 3
Table VII. Dental changes
C A B AvB
Mean Mean Mean Mean diff
Mx 1-palatal plane 0.24 1.2 1.16 0.04 NS
Mx 6-palatal plane 0.93 1.65 1.64 0.01 NS
Mx 1-NA (mm) 0.93 1.23 0.91 0.32 NS
Mx 1-NA angle 6.16 0.15 2.71 3.20 NS
*P ⬍ .01; NS, not significant; diff, difference; SE, standard error.
A-point, ⫹2.63 and ⫹2.43 mm; maxilla to cranial base,
⫹1.67 and ⫹1.41 mm).
Many investigators including Nartallo-Turley and
Turley19 (SNA angle, ⫹2.35°; A-point, ⫹3.34 mm;
maxilla to cranial base, ⫹2.35 mm), Baccetti et al32
(ANS and A-point moved anteriorly 1.7 and 2.0 mm,
respectively) and Macdonald et al26 (SNA angle,
⫹2.59°; nasion to A-point, ⫹1.61 mm; A-point, ⫹2.20
mm) reported similar findings about overall treatment
results from maxillary protraction. In contrast, Saadia
and Torres,36 Gallagher et al,21 Mermigos et al,38 and
Baik8 reported less anterior movement of the maxilla,
whereas Kapust et al10 (SNA angle, ⫹3.71°; nasion to
A-point, ⫹3.30 mm; A-point, ⫹3.93 mm; maxilla to
cranial base, ⫹3.39 mm) reported greater anterior
movement. Although these studies support the efficacy
of maxillary protraction, the differences in results can
be attributed to many factors, including differences in
study design (in our study, we subtracted normal Class
III growth from our results to accurately evaluate the
true maxillary protraction), appliance design, treatment
duration, treatment goals, and the amount of facemask
wear.
Some investigators11,14,15,37-39 have reported simi-
lar overall findings about vertical maxillary movement.
Nartallo-Turley and Turley,19 Kapust et al,10 Baccetti et
al,31 and Macdonald et al26 observed statistically sig-
nificant counterclockwise rotation of the palatal plane,
as we did. The rotation of the palatal plane described in
this and other studies could be affected by many
factors, including site of force application, direction of
elastic traction, and patients’ facial patterns. All pa-
tients in this study were treated with elastic traction
attached mesial to the canine, with the direction of pull
downward (15° to 30°) from the horizontal. Tanne et
al40 and Hata et al9 demonstrated that palatal plane
rotation occurs when the PNS drops more than the
ANS, despite a downward force vector, because the line
of force is directed below the center of resistance of the
maxilla, creating a moment for rotation. The described
movement of the maxilla and associated downward and
backward rotation of the mandible are ideally suited for
Vaughn et al 307
CvA CvB
SE Mean diff SE Mean diff SE
0.57 0.96 NS 0.61 0.92 NS 0.6
0.39 0.72 NS 0.42 0.71 NS 0.42
0.39 0.31 NS 0.42 ⫺0.02 NS 0.41
1.94 ⫺6.80* 1.93 ⫺3.59 NS 1.86
patients with excessive overbites and vertical maxillary
deficiencies. However, subjects with skeletal open bites
were excluded from this study. Caution should be used
for facemask therapy on Class III patients with an
open-bite tendency.
Significant changes in mandibular position also
contributed to the Class III correction. Downward and
backward movement of the chin (SNB angle, ⫺0.86°
and ⫺1.23°; nasion-Pog point, ⫺2.15 and ⫺2.59 mm;
B-point, ⫺1.4 and ⫺2.15 mm; mandible to cranial
base, ⫹2.34 and ⫹3.01 mm) is consistent with the
findings of Ishii et al37 and Takada et al39 with
maxillary protraction, and to those of Nartallo-Turley
and Turley,19 Ngan et al,14 Kapust et al,10 and Mac-
donald et al26 involving palatal expansion with face-
masks. In contrast, Mermigos et al,38 who used maxil-
lary protraction without maxillary expansion, reported
that the mandibular plane remained unchanged. Man-
dibular rotation might be due to a combination of
vertical maxillary movement and eruption of the max-
illary molars, and a distal or retraction force on the
chin.
In addition to skeletal changes in the maxilla and
mandible, orthodontic effects included forward and
downward movement of the maxillary dentition (Table
VII). Ngan et al,14 Baik,8 and Kapust et al10 reported
similar findings. Realizing the limitations of a cranial-
base superimposition to assess orthodontic changes, we
used the occlusal-plane analysis to provide a more
accurate accounting of maxillary tooth movement in-
dependent of osseous changes. Superimposition on the
maxilla confirmed that the maxillary molar moved
forward (⫹1.62 and ⫹2.34 mm). Nartallo-Turley and
Turley19 (⫹1.70 mm) and Ngan et al14 observed similar
maxillary molar movement; Kapust et al10 (2.44 mm)
observed slightly greater maxillary molar movement
using the same method. When orthopedic versus or-
thodontic effects were evaluated (maxillary orthopedic
movement compared with maxillary molar movement),
the occlusal-plane analysis showed an overall ratio of
1:1, which is slightly less than Nartallo-Turley and
Turley (1.4:1) and Kapust et al (1.38:1) found, with
308 Vaughn et al
Table VIII. Soft tissue changes
C A B AvB
Mean Mean Mean Mean diff
Nasiolabial angle ⫺3.2 1.64 5.63 3.99 NS
Cant of upper lip 0.789 1.98 0.24 1.73 NS
ST upper FH % ⫺0.07 0.6 0.46 0.14 NS
ST lower FH % ⫺0.47 0.86 1.66 0.80 NS
*P ⬍ .05; **P ⬍ .01; NS, not significant; diff, difference; SE, standard error.
both measuring greater maxillary orthopedic move-
ment.
Various soft tissue changes combined to improve
the Class III profile (Table VIII). Forward movement of
the upper lip and retraction of the lower lip, coupled
with soft tissue pogonion moving back and menton
moving down, contributed to the more convex profile.
Nartallo-Turley and Turley19 and Kapust et al10 also
reported these changes. The observed soft tissue effects
appeared to result from the induced skeletal changes.
This is consistent with observations by Ngan et al15 that
significant correlations were found between changes of
the sagittal relationships of skeletal and soft tissue
profiles in both the maxilla and the mandible.
To date, this is the only study to evaluate the
treatment effects of protraction facemask therapy, with
or without palatal expansion, in a controlled, random-
ized clinical trial. Kim et al,41 in his meta-analysis of
the effectiveness of protraction facemasks (based on 14
published articles), concluded that there were no dis-
tinct differences between treatments, with or without
palatal expansion, except for maxillary incisor angula-
tion, which increased 2.81° in the nonexpansion treat-
ment group. In our study, no significant difference was
found between the 2 groups in incisor angulation.
Although our results agree with those of Kim et al,41
our study provides the first conclusive evidence to
support that statement. This study also is one of the few
to use a Class III control group, enabling us to measure
the treatment effects beyond normal, measured Class
III growth.
CONCLUSIONS
This study demonstrated that facemask therapy
with or without palatal expansion produced equivalent
changes in the dentofacial complex that combined to
improve the Class III malocclusion. Correction results
from a combination of skeletal and dental changes that
produces an improvement in the skeletal, dental, and
soft tissue relationships. The treated group (n ⫽ 43) had
statistically significant hard and soft tissue movements
affecting the entire dentofacial complex. Skeletal
American Journal of Orthodontics and Dentofacial Orthopedics
September 2005
CvA CvB
SE Mean diff SE Mean diff SE
2.12 4.85* 2.27 8.83** 2.25
1.58 1.19 NS 1.69 -0.54 NS 1.68
0.53 0.67 NS 0.57 0.53 NS 0.56
0.78 1.33 NS 0.83 2.13* 0.82
change was a combination of anterior and vertical
movement of the maxilla, and posterior and downward
movement of the mandible. Dental changes also con-
tributed to the correction, and soft tissue changes
resulted in a more convex profile. These results suggest
that the indication for palatal expansion should be
based on clinical criteria other than assisting the Class
III correction.
We thank Dr Jeff Gornbein and Lena Ting for their
assistance with the statistical analysis and Dr Summer
Blake for her assistance with data entry
REFERENCES
1. McNamara JA Jr. An orthopaedic approach to the treatment of
Class III malocclusion in growing children. J Clin Orthod
1987;21:598-608.
2. Haas AJ. Palatal expansion: just the beginning of dentofacial
orthopedics. Am J Orthod 1970;57:219-55.
3. Haas AJ. Rapid palatal expansion: a recommended prerequisite
to Class III treatment. Trans Eur Orthod Soc 1973;311-8.
4. Williams S, Anderson CE. The morphology of the potential
Class III skeletal pattern in the growing child. Am J Orthod
1986;89:302-11.
5. Hopkins GB. Cranio-facial pattern in mesio-occlusion. Ned
Vereiniging Orthod Studie 1965;81-105.
6. Turley PK. Orthopedic correction of Class III malocclusion with
palatal expansion and custom protraction headgear. J Clin Orthod
1988;22:314-25.
7. Jacobson A, Evans WB, Preston CB, Sadowsky PL. Mandibular
prognathism. Am J Orthod 1974;66:140-71.
8. Baik HS. Clinical results of the maxillary protraction in Korean
children. Am J Orthod Dentofacial Orthop 1995;108:583-92.
9. Hata S, Itoh T, Nakagawa M, Kamogashi K, Ichikawa K,
Matsumoto M, et al. Biomechanical effects of maxillary protrac-
tion on the craniofacial complex. Am J Orthod 1987;91:305-11.
10. Kapust AJ, Sinclair PM, Turley PK. Cephalometric effects of
face mask/expansion therapy in Class III children: a comparison
of three age groups. Am J Orthod Dentofacial Orthop 1998;113:
204-12.
11. Nanda R. Biomechnical and clinical consideration of a modified
protraction headgear. Am J Orthod 1980;78:125-39.
12. Ngan PW, Hägg U, Yiu C, Merwin D, Wei Shy. Treatment
response to maxillary expansion and protraction. Eur J Orthod
1996;18:151-68.
13. Ngan PW, Hägg U, Yiu C, Wei Shy. Treatment response and
long-term dentofacial adaptations to maxillary expansion and
protraction. Semin Orthod 1997;3:255-64.
American Journal of Orthodontics and Dentofacial Orthopedics
Volume 128, Number 3
14. Ngan P, Wei SH, Hägg U, Yui KY, Merwin D, Stickel B. Effect
of protraction headgear on Class III malocclusion. Quintessence
Int 1992;23:197-207.
15. Ngan P, Hägg U, Yui C, Merwin D, Wei SH. Soft tissue and
dentoskeletal profile changes associated with maxillary expan-
sion and protraction headgear treatment. Am J Orthod Dentofa-
cial Orthop 1996;109:38-49.
16. Tanne K, Sakuda M. Biomechanical and clinical changes of the
craniofacial complex from orthopedic maxillary protraction.
Angle Orthod 1991;61:145-52.
17. Turley PK. Orthopedic correction of the Class III malocclusion:
retention and phase II therapy. J Clin Orthod 1996;30:313-24.
18. Williams MD, Sarver DM, Sadowsky LP, Bradley E. Combined
rapid palatal expansion and protraction facemask in the treatment
of Class III malocclusions in growing children: a prospective
long-term study. Semin Orthod 1997;3-4:265-74.
19. Nartallo-Turley PE, Turley PK. Cephalometric effects of com-
bined palatal expansion and facemask therapy on Class III
malocclusion. Angle Orthod 1998;68:217-24.
20. Merwin D, Ngan P, Hägg U, Yiu C, Wei Shy. Timing for
effective application of anteriorly directed orthopedic force to the
maxilla. Am J Orthod Dentofacial Orthop 1998;112:292-9.
21. Gallagher RW, Miranda F, Buschang PH. Maxillary protraction:
treatment and posttreatment efects. Am J Orthod Dentofaial
Orthop 1998;113:612-9.
22. Baccetti T, McGill JS, Franchi L, McNamara JA Jr, Tollaro I.
Skeletal effects of early treatment of Class III malocclusion with
maxillary expansion and face-mask therapy. Am J Orthod
Dentofacial Orthop 1998;113:333-43.
23. Silva Filho OG, Magro AC, Filho LC. Early treatment of the
Class III malocclusion with rapid maxillary expansion and
maxillary protraction. Am J Orthod Dentofacial Orthop 1998;
113:196-203.
24. Suda N, Ishii-Suzuki M, Hirose K, Hiyama S, Suzuki S, Kuroda
T. Effective treatment plan for maxillary protraction: is the bone
age useful to determine the treatment plan? Am J Orthod
Dentofacial Orthop 2000;118:55-62.
25. Gu Y, Rabie A, Hägg U. Treatment effects of simple fixed
appliance and reverse headgear in correction of anterior cross-
bites. Am J Orthod Dentofacial Orthop 2000;117:691-9.
26. Macdonald KE, Kapust AJ, Turley PK. Cephalometric changes
after the correction of Class III malocclusion with maxillary
expansion/facemask therapy. Am J Orthod Dentofacial Orthop
1999;116:13-24.
Vaughn et al 309
27. Kilicoglu H, Kirlic Y. Profile changes in patients with Class III
malocclusions after Delaire mask therapy. Am J Orthod Dento-
facial Orthop 1998;113:453-62.
28. Davis WM, Kronman JH. Anatomic changes induced by splitting
of the midpalatal suture. Angle Orthod 1969;39:126-32.
29. Haas AJ. Rapid expansion of the maxillary dental arch and nasal
cavity by opening the midpalatal suture. Angle Orthod 1961;31:
73-90.
30. Wertz RA. Skeletal and dental changes accompanying rapid
midpalatal suture opening. Am J Orthod 1970;58:41-66.
31. Baccetti T, Franchi L, McNamara JA. Treatment and posttreat-
ment craniofacial changes after rapid maxillary expansion and
facemask therapy. Am J Orthod Dentofacial Orthop 2000;
118.404-13.
32. Dahlberg AG. Statistical methods for medical and biological
students. New York: Interscience Publications; 1940.
33. Johnston LE. A comparative analysis of Class II treatments,
monograph 19. Craniofacial Growth Series. Ann Arbor: Center
for Human Growth and Development; University of Michigan;
1986. p. 103-48.
34. Oppenheim A. A possibility for physiologic orthodontic move-
ment. Am J Orthod Oral Surg 1944;30:345-92.
35. Baik HS. Assessment of skeletal and dental changes by maxillary
protraction. Am J Orthod Dentofacial Orthop 1998;114:492-502.
36. Saadia M, Torres E. Sagittal changes after maxillary protraction
with expansion in Class III patients in the primary, mixed, and
late mixed dentitions: a longitudinal retrospective study. Am J
Orthod Dentofacial Orthop. 2000;117:669-80.
37. Ishii H, Morita S, Takeuchi Y, Nakamura S. Treatment effect of
combined maxillary protraction and chin cup appliance in severe
Class III cases. Am J Orthod Dentofacial Orthop 1987;92:304-
12.
38. Mermigos J, Full CA, Andreasen G. Protraction of the maxillo-
facial complex. Am J Orthod Dentofacial Orthop 1990;98:47-55.
39. Takada K, Petdachai S, Sakuda M. Changes in dentofacial
morphology in skeletal Classs III children treated by a modified
maxillary protraction headgear and a chin cup. Eur J Orthod
1993;15:211-21.
40. Tanne K, Hiraga J, Kakiuchi K, Yamagata Y, Sakuda M.
Biomechanical effect of anterior directed extraoral forces on the
craniofacial complex: a study using finite element method. Am J
Orthod Dentofacial Orthop 1989;95:200-7.
41. Kim JH, Viana M, Graber TM, Omerza F, BeGole E. The
effectiveness of protraction facemask therapy: a meta-analysis.
Am J Orthod Dentofacial Orthop 1999;115:675-85.