European Journal of Orthodontics 32 (2010) 171–176
Advance Access Publication 3 December 2009
© The Author 2009. Published by Oxford University Press on behalf of the European Orthodontic Society.
All rights reserved. For permissions, please email: email@example.com.
Transpalatal arches are routinely used in orthodontic
treatment in both the permanent and mixed dentition. Their
mode of action can be divided into passive, to stabilize or
reinforce anchorage, or active, to enable tooth movement.
Thus, tooth movement may be undertaken for a single tooth
or blocks of teeth in the horizontal, sagittal, and vertical
directions. A number of active tooth movements are possible
with a palatal arch, including derotation of unilateral or
bilateral rotated molars (Cooke and Wreakes, 1978; Ten
Hoeve, 1985; Dahlquist et al., 1996; Ingervall et al., 1996).
Transpalatal arches may also be used to correct molar
crossbites, which is well described in a prospective clinical
study by Ingervall et al. (1995). Further reports in the
literature describe the use of palatal arches as a mode for
asymmetric or symmetric distalization (Ten Hoeve, 1985;
Mandurino and Balducci, 2001) and buccal or lingual root
torque of the upper molars (Baldini and Luder, 1982).
More commonly, palatal arches are used to reinforce
anchorage and prevent mesial movement of the upper first
permanent molars during treatment. The anchorage value
is increased by maintaining a fixed intermolar width
across the arch, so that as the molars loose anchorage by
drifting forwards, their roots engage the buccal cortex,
A randomized clinical trial to compare the Goshgarian and
Nance palatal arch
N. Stivaros*, C. Lowe**, N. Dandy**, B. Doherty* and N. A. Mandall*
*School of Dentistry, University of Manchester and **Hightown Orthodontic Practice, Crewe, UK
SUMMARY The aim of this trial was to evaluate whether a Nance or Goshgarian palatal arch was most
effective for prevention of mesial drift, distal tipping, prevention of mesio-palatal rotation of the upper
first permanent molars, and patient comfort and ease of removal. Patients were recruited from a district
general hospital and a specialist orthodontic practice and randomly allocated to a Goshgarian (n = 29)
or a Nance (n = 28) group. Pre-treatment study models (T1) were taken followed by the placement of the
palatal arch, premolar extractions, and upper and lower fixed appliances. The clinical end point was 6
months (T2), at which time, an impression for an upper study model was taken. The amount of upper first
permanent molar mesial movement, distal tipping, and mesio-palatal rotation was measured by scanning
T1 and T2 study models and then using a software program to calculate molar changes. In addition, the
patients recorded their discomfort scores using a seven-point Likert scale at each recall visit.
Forty-nine patients (86 per cent) completed the trial. t-tests were used to compare molar movements
between the Goshgarian and Nance palatal arch groups. There were no statistically significant differences
between the palatal arches in terms of prevention of mesial drift or distal tipping (P > 0.05). There was
a statistically significant difference in the amount of molar rotation between the arch types, with both
exhibiting some disto-palatal rotation even though they were not activated for this movement. The
Goshgarian palatal arch produced marginally more disto-palatal rotation than the Nance arch (P = 0.02),
although this may not be considered clinically significant. A Mann–Whitney test revealed that there was
also a statistically significant difference in pain scores between the Goshgarian and the Nance arch, with
the latter being associated with more discomfort (P = 0.001). This trial did not support any preference in
the use of the Goshgarian or Nance palatal arch, unless the slightly reduced patient discomfort with the
Goshgarian arch is considered significant.
which theoretically will prevent further forward drift.
However, this concept of cortical anchorage is not
supported scientifically; joining the molar teeth together,
thus doubling their root surface area and therefore
increasing their resistance to unwanted mesial drift; or
controlling molar rotation and tipping and thus, to some
extent, restricting forward movement of the upper first
The Nance (Nance, 1947) and Goshgarian (Goshgarian,
1972) palatal arches (Figure 1) have been described in the
literature as providing reinforcement of anchorage, but no
comparison of the effectiveness of two types of palatal arch
have been scientifically evaluated.
Therefore, the aim of the study was to evaluate whether a
Nance or a Goshgarian palatal arch was the most effective
for prevention of mesial drift, distal tipping, prevention of
mesio-palatal rotation of the upper first permanent molars,
and patient comfort and ease of removal.
Subjects and method
Approval for the study was obtained from North Manchester
(03/NM/626) and Cheshire (M248/03) Local Regional
N. STIvAROS ET AL.
Sample size calculation
Since there was no comparative literature to suggest an
expected difference between the Nance and Goshgarian
palatal arches for loss of anchorage, a clinically significant
difference of 2 mm of mesial movement between groups
was used. A sample size in each group of n = 23, with a
0.050 two-sided significance level, will have a 90 per cent
power to detect a difference of 2 mm of mesial movement
of the upper first permanent molars between the Nance and
the Goshgarian palatal arch groups, assuming a common
standard deviation of 2 mm. Thus, a total of 46 patients
needed to be recruited.
The sample was obtained by approaching consecutive
patients, who fulfilled the inclusion criteria, attending for
fixed orthodontic treatment at Tameside General Hospital,
Ashton-under-Lyne, Lancashire, UK and a specialist
orthodontic practice in Crewe, Cheshire, UK. The treatment
was conducted by one orthodontic consultant (NAM), one
specialist practitioner in orthodontics (CL), and one
orthodontic postgraduate (NS).
Figure 1 The Nance (a) and Goshgarian (b) palatal arches.
1. Patients aged 10-17 years at the start of treatment.
2. Upper premolar extractions.
3. Patient just about to commence orthodontic treatment.
4. Upper and/or lower preadjusted edgewise appliance
(McLaughlin, Bennett, Trevisi/MBT prescription) and
stainless steel brackets.
5. Patient and parent informed and written consent.
1. Requiring upper first permanent molar extractions.
2. Correction of a crossbite with expansion of the upper arch.
3. Non-extraction upper arch treatment.
4. Cases that required extra oral anchorage reinforcement
or distal movement of the upper first permanent molars.
5. Patients requiring orthognathic surgery.
6. Lack of consent.
Randomization was carried out using random number
tables, where even numbers were assigned to the Nance
palatal arch group and odd numbers to the Goshgarian
palatal arch group. A restricted randomization method was
used in blocks of 10 to ensure that equal numbers of patients
were allocated to each treatment group. The treatment
allocation was concealed in an opaque envelope and labelled
with the study identification number. When the patient
consented to the trial, the envelope was opened to reveal the
treatment allocation. During the trial, the operator could not
be blind to treatment allocation; however, the examiner who
measured the study models (NS) did not know whether a
Nance or a Goshgarian palatal arch had been used.
The type of malocclusion was recorded using the Incisor
Classification of the British Standards Institution (1983).
A standardized procedure was then used for every patient
and the palatal arches fitted immediately prior to the upper
arch extractions and fixed appliance placement. Both
palatal arches were constructed using 0.9 mm stainless
steel and were soldered to the molar bands. In the case of
the Goshgarian palatal arch, the loop faced distally and
was constructed so that it was 2–3 mm away from the
palatal mucosa. The Nance palatal arch was designed with
a large acrylic button extended to cover the steepest and
deepest depth of the anterior palatal vault. Both palatal
arches were cemented with chemical cure glass ionomer
cement (Intact; Orthocare, Bradford, West Yorkshire, UK).
The aim was not to activate the palatal arch to derotate the
molars since the effectiveness of the prevention of mesio-
palatal rotation was being investigated. In two subjects
(one in each group) where the upper first molars were
considerably mesio-palatally rotated at the start of
treatment, a derotation activation was placed bilaterally in
the palatal arch and these patients were excluded from the
A CLINICAL COMPARISON OF TWO PALATAL ARCHES
The remaining fixed appliance was then fitted using MBT
prescription with an archwire sequence of 0.016 inch nickel
titanium (Ni-Ti), 0.018 × 0.025 inch Ni-Ti and 0.019 ×
0.025 inch stainless steel. During the levelling and aligning
phase of treatment, which comprised a 6 month observation
period, no mechanics were used that may have brought the
upper first permanent molars more mesially than would be
assumed by loss of anchorage alone. In addition, no canine
lacebacks were placed. The palatal arch was removed at the
clinical end point of 6 months (T2). The patients were then
sent away for 1 week and told to brush their palate. This was
to ensure that the palatal mucosa in the Nance palatal arch
group was not swollen or hyperplastic when the impression
was taken, as this would have enabled the observer to
determine that the patient had worn the Nance palatal arch.
One week after removal of the palatal arch, an end point
upper alginate impression was taken and either new upper
first molar bands were cemented or the palatal arch was
re-cemented if it was deemed clinically necessary.
The following outcome measures were assessed in relation to
the position of the upper first permanent molars: mesial
movement, distal tipping, prevention of mesio-palatal rotation,
patient discomfort, ease of removal of the palatal arch, and
whether or not local anaesthesia was required for removal.
Mesial movement, distal tipping, and prevention of mesio-
palatal rotation were assessed by comparing the difference in
the position of the upper first permanent molars on the T1 and
T2 study models. The models were scanned with an updated
version of the Konica Minolta vivid 700 three-dimensional
(3D) laser scanner (www.konicaminolta.com). The laser
digitizer is placed on a moveable turntable and a point of
laser light is shone on the surface of the object to be scanned.
The reflected beam is then intercepted by a sensor, which
converts it into 3D co-ordinates; this is known as triangulation.
The computer software program, Rapid Form 2004 (Konica),
was then used to create the 3D model and calculate the tooth
movements of interest using the palatal rugae as reference
points. Figure 2 shows the measurements taken on one set of
start (T1) study models.
Patient discomfort was measured on a seven-point Likert
scale, at each visit, where a score of 1 indicated no pain and
a score of 7 severe pain. A mean score was calculated. Ease
of removal of each palatal arch was recorded as whether the
palatal arch embedded or impinged on the palatal mucosa
(yes/no) and whether local anaesthesia was required for its
Systematic error was reduced by the examiner being blind
to the treatment group when the model measurements were
carried out. The examiner was also blind as to whether the
models were obtained at T1 or T2. All models were measured
in a random order so the patients’ T1 and T2 models were
not measured consecutively. Random error was reduced by
measuring all study models twice and calculating a mean.
Examiner calibration and reliability
The examiner was trained in the use of the laser scanner
and calibration was carried out using four plaster calibration
cubes with different known heights and widths that had
been verified by the Engineering Department, University
of Manchester, UK as the gold standard. Intra-examiner
reliability of the measurements was assessed by
re-measuring 20 study models after an interval of at least
The data were checked for normality and simple summary
statistics produced. t-tests were used to compare the palatal
arch groups in terms of upper first permanent molar mesial
movement, distal tipping, and prevention of mesio-palatal
rotation. Left and right molar movements were averaged
for analysis. A Mann–Whitney test was used to compare
pain scores between the two palatal arch groups and chi-
square statistics to compare the groups in terms of ease of
removal of the palatal arch and whether or not local
anaesthesia was required. Examiner calibration and
reliability was assessed using intraclass correlation
coefficients (ICC). Pearson correlation coefficients for
examiner calibration were 1.00 for height measurements
and 0.99 for width measurements. Intra-examiner
reliability for all tooth movement measurements was high,
with an ICC of 0.98 or above.
The trial profile is shown in Table 1. Of the 57 patients
initially registered, the final sample size for data analysis
was 49. Six patients (four females and two males) were not
included in the data analysis since they were lost to
follow-up. Two further patients needed disto-palatal
activation of their palatal arches so were excluded from the
analysis. The mean age for the Goshgarian palatal arch
group was 14 years 6 months and that of the Nance group
14 years 3 months. There was no statistically significant
difference between the groups for age (P = 0.54), thus
exhibiting pre-treatment equivalence for this variable. In
addition, there was no statistically significant difference in
gender between the two palatal arch groups (P = 0.035) or
initial malocclusion (P = 0.29).
Upper first permanent molar mesial movement, distal
tipping, and prevention of mesio-palatal rotation
Table 2 shows that there was no statistically significant
difference between the Goshgarian and Nance palatal arch in
N. STIvAROS ET AL.
Figure 2 Upper first permanent molar mesial movement (a and b), tipping (c), and rotation (e and d).
terms of prevention of mesial movement or distal tipping (P >
0.05). The average mesial movement for both groups was in
the order of 0.5–1 mm over the 6 month period. Overall, there
was, in fact, mesial tipping for both groups and this was in the
order of 2–3 degrees, with wide standard deviations.
In contrast, there was a statistically significant difference
in the prevention of mesio-palatal rotation between the palatal
arch groups. The Goshgarian palatal arch group exhibited a
small amount of disto-palatal rotation of around 4.5 degrees
compared with the Nance at just over 2 degrees (P = 0.017).
A CLINICAL COMPARISON OF TWO PALATAL ARCHES
Table 1 Trial profile.
Registered patients n = 57
Not randomized n = 0 Randomization n = 57
Goshgarian palatal arch
Intervention n = 29
Followed-up n = 24
Withdrawn n = 1*
Intervention ineffective n = 0
Lost to follow-up n = 4
Completed trial n = 24
Nance palatal arch
Intervention n = 28
Followed-up n = 25
Withdrawn n = 1*
Intervention ineffective n = 0
Lost to follow-up n = 2
Completed trial n = 25
*Withdrawn from analysis because palatal arch activated for derotation.
Table 2 Comparison of Nance and Goshgarian palatal arches for upper first permanent molar mesial movement, tipping, and rotation.
Tooth movement Palatal archMean (SD)SE mean
t valueP value
95% Confidence interval
Mesial movement (mm)Goshgarian
0.68 0.50−0.44 to 0.90
Rotation * (°)2.500.02 0.44 to 4.19
Tipping * (°)0.370.72 −3.07 to 4.41
*+ve value indicates disto-palatal rotation or distal tipping; −ve value indicates mesio-palatal rotation or mesial tipping.
This was surprising as the palatal arches were not activated
for disto-palatal rotation but there was still a tendency for this
movement to occur between 2–4 degrees in both groups.
Patient discomfort scores and ease of removal of the palatal
There was a statistically significant difference in discomfort
scores between the two groups, with the Nance palatal arch
reported to be more uncomfortable (median Likert score = 2)
compared with the Goshgarian (median Likert score = 1;
P = 0.001). No gagging problems were reported with either
palatal arch design. Two Nance palatal arches, but no
Goshgarian palatal arches, were embedded in the mucosa (chi-
square 2.002, 1 degree of freedom, P value = 0.16). No local
anaesthesia was required to remove any of the palatal arches.
Assessment of potential bias through patients who dropped
out of the study was carried out by comparing the malocclusion,
gender, and age between the patients staying in and those
dropping out of the study. There was no statistically significant
difference between the patients remaining in the study and
those who did not (P > 0.05), so there should be no bias as a
result of losing eight patients from the sample.
This trial did not find a clinically significant difference
between the Nance and Goshgarian palatal arches in terms
of prevention of mesial drift, distal tipping, or prevention of
mesio-palatal rotation of the upper first permanent molars.
However, there was a difference between the two groups in
terms of increased discomfort experienced by the patients
wearing the Nance palatal arch.
The results of this study revealed that the amount and
type of tooth movement from the two appliances was small.
When the average molar movement between the right and
left sides was measured for each palatal arch type, the
Goshgarian palatal arch allowed the molars to mesialize
slightly more than the Nance (0.94 versus 0.72 mm) over a
6 month period. These values were very small and indeed
there was no statistically or clinically significant difference
between the arch types in preventing anchorage loss. It is
difficult to make comparison with previous literature
because of the lack of randomized clinical trials. Although
not directly related to the appliances used in this trial,
Rebellato et al. (1997) conducted a prospective clinical trial
investigating whether a lower lingual arch was effective at
preventing mesial migration of the lower first permanent
molars. Their results showed that the molars mesialized in
the treatment group by 0.29 mm, which was marginally less
than the findings in the present study. However, caution
must be exercised when making comparisons between
studies: firstly the investigation by Rebellato et al. (1997)
had no fixed appliance in situ and secondly they were
investigating mandibular molar movement, which tends to
be slower than its maxillary counterpart.
This trial demonstrated that in some patients the molars
rotated mesio-palatally during treatment but the overall
average was for the molar to derotate disto-palatally (4.4
degrees with the Goshgarian and 2.1 degrees with the
Nance). The difference was statistically significant;
however, it is doubtful whether a difference of only 2.3
degrees between groups is clinically significant. In addition,
it is difficult to explain why any disto-palatal rotation
occurred in either group, as the palatal arches were not
activated to straighten the molars. It is possible that the
0.019 × 0.025 inch stainless steel working archwire, in
conjunction with the appliance prescription, caused a mean
disto-palatal molar rotation, despite a rigid soldered palatal
arch being in place.
N. STIvAROS ET AL.
Although the findings demonstrated that in some patients
the molars tipped distally, this was in a minority of cases
and the overall tendency was for the molars to tip mesially
(2.09 degrees with the Goshgarian and 2.75 degrees with
the Nance). The difference between the groups was not
statistically significant and thus one appliance was no better
than the other for preventing mesial tipping. It was perhaps
surprising that more distal tipping was not observed in the
Goshgarian group as Chiba et al. (2003) suggested that
tongue pressure on a distally directed U loop would tip the
molars distally. It is probable that the design in this study to
minimize patient discomfort, meant that the position of the
U loop was too near the palate for tongue pressure to have
had a significant tipping effect.
The clinical end point of 6 months was used to avoid any
delay in treatment since some patients were ready for space
closure mechanics. It would have been clinically useful to
continue measuring anchorage loss during space closure.
However, this was not carried out because some patients
required no further anchorage reinforcement and the palatal
arch needed to be removed to allow mesial movement of the
upper buccal segments.
When the possibility of the palatal arch embedding in the
oral mucosa was considered, again there were no statistically
significant differences between the palatal arch types. This is
not in agreement with the general clinical observation that
Nance palatal arches tend to embed into the palatal mucosa
because of the proximity of the palatal acrylic to the palate.
However, the two arches that did embed were of the Nance
design but local anaesthesia was not needed for their removal.
It may be that the absence of acrylic embedding into the
palatal mucosa was because the surface area of the acrylic
was large, or that no significant space closure mechanics
were used during the 6 month observation period.
Generally, no differences were observed between the clinical
behaviour of the Goshgarian and Nance palatal arches.
Although some statistically significant differences were found
between the groups, examination of the mean values did not
suggest a clinically meaningful difference. Thus, when a
clinician is choosing which palatal arch to use, both designs are
effective in only allowing around 1 mm of mesial movement of
the upper first permanent molars over a 6 month period.
Consideration was given to whether to include a group
with no palatal arch. However, there was a possibility that
a group with low anchorage requirement, and no need for
a palatal arch, may have been inherently different compared
with patients requiring one. Therefore, a group with no
palatal arch could have had different start characteristics
and not be comparable with the other groups at baseline.
For this reason, it was decided to compare two types of
palatal arch and not attempt to introduce a control group.
There was no statistical or clinical advantage in the use of
either the Goshgarian or Nance palatal arch in terms of
prevention of mesial molar movement or distal molar
tipping during the first 6 months of orthodontic treatment.
Differences in prevention of mesio-palatal rotation are
unlikely to be clinically significant. The trial does not
support the use of one palatal arch or another, unless the
slightly reduced patient discomfort for the Goshgarian arch
is considered clinically significant.
Address for correspondence
Dr N A Mandall
Tameside General Hospital
Lancashire OL6 9RW
Baldini G, Luder H 1982 Influence of arch shape on the transverse
effects of transpalatal arches of the Goshgarian type during
application of buccal root torque. American Journal of Orthodontics
British Standards Institution 1983 British Standard Incisor Classification.
Glossary of dental terms BS4492. British Standard Institute, London
Chiba Y, Motoyoshi M, Namura S 2003 Tongue pressure on loop
transpalatal arch during deglutition. American Journal of Orthodontics
and Dentofacial Orthopedics 123: 29–34
Cooke M S, Wreakes G 1978 Molar derotation with a modified palatal
arch: an improved technique. British Journal of Orthodontics 5:
Dahlquist A, Gebauer U, Ingervall B 1996 The effect of a transpalatal arch
for the correction of first molar rotation. European Journal of
Orthodontics 18: 257–267
Goshgarian R A 1972 Orthodontic palatal arch wires. United States
Government Patent Office. Alexandria, Virginia (Patent number 3792529)
Ingervall B, Göllner P, Gebauer U, Fröhlich K 1995 A clinical investigation
of the correction of unilateral first molar crossbite with a transpalatal
arch. American Journal of Orthodontics and Dentofacial Orthopedics
Ingervall B, Hönigl K D, Bantleon H P 1996 Moments and forces delivered
by transpalatal arches for symmetrical first molar rotation. European
Journal of Orthodontics 18: 131–139
Mandurino M, Balducci L 2001 Asymmetric distalisation with a TMA
transpalatal arch. Journal of Clinical Orthodontics 35: 174–178
Nance H N 1947 The limitations of orthodontic treatment 1: mixed
dentition diagnosis and treatment. American Journal of Orthodontics
and Oral Surgery 33: 177–223
Rebellato J, Lindauer S J, Rubenstein L K, Isaacson R J, Davidovitch M,
vroom K 1997 Lower arch perimeter preservation using the lingual
arch. American Journal of Orthodontics and Dentofacial Orthopedics
Ten Hoeve A 1985 Palatal bar and lip bumper in non-extraction treatment.
Journal of Clinical Orthodontics 4: 272–291