ORIGINAL ARTICLE

Clinical comparison and performance perspective of

three aligning arch wires

T. Justin W. Evans, BDS, MScD, MOrth, FDSRCS,a

Malcolm L. Jones, BDS, MSc, PhD, DOrth, FDSRCS,b and
Robert G. Newcombe, MA, PhD
Cardiff, Wales, U. K.

Objectives: To clinically evaluate three commonly used orthodontic tooth aligning arch wires: 016 3
022 inch active martensitic medium force nickel titanium, 016 3 022 inch graded force active
martensitic nickel titanium, and 0.0155 inch multistrand stainless steel.
Design: A prospective randomized clinical trial.
Data source: Measured serial study casts of dental arches for 112 assigned arch wires from 56
consecutive patients. Analysis based on completed records for 98 arch wires and 51 patients.
Method: A consecutive sample of 56 patients requiring both upper and lower fixed appliance
therapy were randomly allocated two different arch wires from a possible three under trial. Good
quality impressions were taken of the dental arches at the designated serial stages of alignment
(start, T0; 4 weeks, T4; 8 weeks, T8). The resultant casts were measured on a Reflex Microscope to
record the change in individual tooth alignment both in three and two dimensions (horizontal plane
only).
Results: The measurement error was within acceptable limits (range, 0.05 to 0.09 mm) and showed
no significant bias. ANOVA statistical models were fitted to the data to adjust for a number of
variables. No significant difference in aligning capability (p . 0.05), in either two or three
dimensions, was demonstrated between the three arch wires in the trial.
Conclusion: Heat activated nickel titanium arch wires failed to demonstrate a better performance
than the cheaper multistrand stainless steel wires in this randomized clinical trial. The failure to
demonstrate in vivo superiority at the clinical level may be due to the confounding effects of large
variations in individual metabolic response. Alternatively, it may be that in routine clinical practice
NiTi-type wires are not sufficiently deformed to allow their full superelastic properties to come in to
play during initial alignment. (Am J Orthod Dentofacial Orthop 1998;114:32-39)

Fixed appliance therapy is a treatment

modality based on the theory that by applying light
continuous force to a tooth, it may be moved
optimally through the alveolar bone of the jaws
without causing permanent damage.1-10 However,
quantifying this force is difficult because of individ-
ual variation in tissue response, root morphology,
Support for this project came from the Welsh Scheme for the Develop-
ment of Health and Social Research.
From the University of Wales College of Medicine.
a
Registrar at Department of Dental Health and Development, University
of Wales College of Medicine. Now Senior Registrar at Westmead Dental
Hospital, NSW, Australia.
b
Professor of Orthodontics and Head of Department of Dental Health and
Development, University of Wales College of Medicine.
c
Senior Lecturer in Medical Statistics, University of Wales College of
Medicine.
Reprint requests to: M. L. Jones, BDS, MSc, PhD, DOrth, FDSRCS,
Department of Dental Health & Development, Dental School, University
of Wales College of Medicine, Heath Park, Cardiff. Wales CF4 4XY.
United Kingdom.
Copyright © 1998 by the American Association of Orthodontists.
0889-5406/98/$5.00 1 0 8/1/85119
32
and the type of tooth movement induced.11-14 In
order for the clinician to choose the most appropri-
ate arch wire during fixed appliance therapy, an
understanding of the optimal characteristics is nec-
essary. In the past these properties have been as-
sessed almost exclusively by ex vivo “bench-testing”
methods. However, it does not necessarily follow
that these properties are mirrored in vivo in the
clinical environment.
In an attempt to provide clinical meaning to
much of the recent ex vivo testing, a number of arch
wire characteristics have been described that are
desirable for optimum performance during treat-
ment: springback, stiffness, formability, modulus of
resilience, biocompatibility, low friction, and join-
ability.9,15-17 If the ideal aligning arch wire is one
capable of producing light and continuous forces
over a prolonged time period, then multistranded
stainless steel, stable and active nickel titanium
alloys would seem to be the closest arch wire
materials to fulfilling the ideal requirements.18 Such
American Journal of Orthodontics and Dentofacial Orthopedics
Volume 114, No. 1
wires are termed “active” because they are subject
to phase transformation under different conditions
of temperature and stress. Essentially the body-
centered cubic austenitic phase predominates under
higher temperatures and reduced stress, whereas
the hexagonal close packed martensitic phase pre-
dominates with decreasing temperature and increas-
ing stress. By varying the temperature during the
manufacturing stage, the relative percentages of
austenite/martensite can be changed to alter the
eventual force delivery/deflection characteristics for
a given diameter of arch wire. It has been recog-
nized that a martensitic wire will deliver 25% to
30% of the force of an austenitic wire initially32 and
with a greater working range such a wire should
confer benefits during the initial tooth alignment
phase of a fixed appliance treatment. The various
theoretical advantages of these different types of
wire have been reviewed previously in more detail
by one of us.18
With an ever-increasing range of alloys, and
alloy specifications on the market, claims of “vastly
superior” wire characteristics should be assessed at
the clinical level. However in vivo interpretation of
aligning properties has been largely limited to anec-
dotal case reports.19-23
In view of the fact that material science has
provided the clinician with aligning and leveling
arch wires possessing superior force/delivery char-
acteristics, one might confidently expect this to be
evident in the rate and amount of tooth alignment at
the clinical level. Although assessment of clinical
behavior can only be determined through con-
trolled, prospective, randomized, clinical trials,
there have only been two published trials of this kind
to date,24-26 and neither has been able to substanti-
ate vastly superior behavior of ‘active’ austenitic
(superelastic) arch wires at the clinical level.
O’Brien et al.24 compared the aligning capabilities
of a superelastic wire versus its stabilized martensitic
variant, whereas West25 compared the clinical effec-
tiveness of a superelastic alloy versus multistranded
stainless steel.
However, neither study assessed the aligning and
leveling capability for greater than 42 days, and
neither study assessed the clinical performance of
active martensitic (heat-activated) arch wires.
The aim of the present study was to evaluate the
aligning and leveling capabilities of three arch wires
currently (in 1996) available,27 over a 4- and 8-week
period, in order to assess their ability to deliver the
clinical advantages claimed by manufacturers based
on data derived from previous bench test analyses.
The three arch wires to be tested clinically were:
● Wire A, 016/022 inch medium force active
martensitic nickel titanium (Titanium Heat
Memory Wire/American Orthodontics)
Evans, Jones, and Newcombe 33
● Wire B, 0155 inch multistranded stainless steel
(Dentaflex/Dentarium)
● Wire C 016/022 inch graded force, active mar-
tensitic nickel titanium (Bioforce Sentalloy/
GAC inc.).
MATERIAL AND METHODS
After a pilot study28 and previous work in this area,25
it was estimated that the inclusion of 100 dental arches in
this trial, in other words 50 subjects (50 upper and 50
lower arches), would provide adequate statistical power
(80%) to detect a significant difference in the perfor-
mance of any two separate arch wires (p , 0.05). A
detailed protocol was designed and approval obtained for
the project from the local ethics committee.
Inclusion criteria
Fifty-six consecutive patients, drawn from two centers,
that required upper and lower arch alignment were in-
cluded in the study. This provided a total of 112 fitted arch
wires. Each patient, from whom informed consent had
been obtained previously, was allocated one upper and
one lower aligning arch wire, each of a different type from
a possible three under investigation; allocation was pre-
determined and randomized.
Patients were eligible to take part in the trial provided
that the following conditions were met:
● Upper and lower fixed appliances were required.
● The patient had not received previous orthodontic
treatment.
● Any initial phase of expansion with a quadhelix had
been stabilized for 2 months.
● Any initial phase of active distal molar movement
had been stable for 2 months.
● The patient was under 18 years of age.
It was felt that all patients who may have experienced
periodontal disease, and hence loss of attachment, also
should be excluded.
Data collection
For each patient a good quality alginate impression of
each dental arch was taken at the prealignment phase (T0)
after placement of bonded attachments and bands. The
allocated arch wire was then ligated by one of six clinicians
as fully as possible into the bracket with the clinicians
preferred method (usually elastomeric rings), and at the
routine follow-up appointment at 4 weeks (T4) a further
alginate impression taken. After full relegation, a further
4 weeks of alignment and leveling was allowed to con-
tinue, and at this stage (T8, 8 weeks into the trial) further
impressions were taken, and the initial aligning arch wire
discarded.
Cast analysis
Alginate impressions were cast with a 50/50 mixture of
plaster and stone. The models were then trimmed for ease
34 Evans, Jones, and Newcombe
Fig. 1. White dots show digitized points from which the
alignment measure was derived. Points on cusps were
used to generate an occlusal plane.
of horizontal orientation to the table of a reflex micro-
scope (Reflex Measurements, Ltd., Butleigh, England).
Once a series of three consecutive casts for a specific arch
had been attained, (T0, T4, and T8), each one was placed
on the moveable Reflex Microscope table. Fiducial points
were noted for each tooth from the mesial aspect of the
first molar to the contralateral first molar, and digitized in
turn (Fig. 1). Posteriorly the external and occlusal line
angle of each bracket/band was used while anteriorly the
incisal edges were used (Fig. 1).
The raw data consisted of measurements of each of
the 11 intertooth (contact) point distances within each
arch; these readings were available at the 0, 4, and 8 week
stage. The changes at each intertooth point distance from
one visit to the next were collected, along with the change
from the first to last visit. Perhaps it should be stressed
that although the points were digitized in three dimen-
sions, the data for the intertooth distance in each instance
gave one value that could then be added to give a score for
each dental arch.
In order to assist in statistical analysis of the raw data
and to fulfill the objectives of the study, data for the
intertooth distances were collated and then summated for
anterior (3-2, 2-1, 1-1, 1-2, 2-3), and posterior (6-5, 5-4,
4-3, 3-4, 4-5, 5-6) sites, allowing analysis of arch segments,
as well as producing a figure for the average alignment
over the whole arch.
In addition to these three-dimensional measurements,
the buccal cusps of the premolars and the mesiobuccal
cusp of the first permanent molar were digitized to
produce two-dimensional alignment readings by projec-
tion to a horizontal plane, reducing the potential source of
vertical measurement error as previously described.29,30
This method of measuring the tooth alignment capabili-
ties of arch wires has been described, in detail, previous-
ly26 and is well established.
Measurement error
In order to assess measurement error, duplicate read-
ings were taken of cast series at the commencement (T0),
American Journal of Orthodontics and Dentofacial Orthopedics
July 1998
Table I. Summary statistics for placement of three arch wires
A B C
Upper Archwires 19 14 16
Lower Archwires 13 17 19
Mean T0-T4 time span (days) 28.4 29.0 28.3
Mean T4-T8 time span (days) 28.7 28.3 29.2
mid point (T4), and end of the trial (T8). This was carried
out to give a value for reproducibility throughout the
whole trial period. The produced mean value for each
variable was used in the analysis of the data.
Statistical analysis
Raw data consisted of measurements of each of the 11
intertooth point distances within each arch from the
mesial of the first molar to the mesial of the contralateral
molar.
For each arch, data were available for two- and
three-dimensional readings, at the 0, 4, and 8 week stage
(T0, T4, T8). The changes at each intertooth point distance
from one visit to the next were designated T0 to T4, and T4
to T8, along with the overall change from T0 to T8.
Changes in intertooth distances were then summated to
represent overall tooth movement for the whole arch, the
anterior segment (mesial of canine to mesial of canine),
and posterior segments (mesial of molar to distal of
canine).
Analysis of variance (ANOVA) models were fitted in
order to assess differences between the three arch wires to
determine whether a significant difference exists in their
ability to bring about rapid alignment. Because of the
number of variables involved for each wire type, the
ANOVA model was fitted to take account of the main
effects for subject, order of wire placement, upper or
lower arch, and arch wire material. This was carried out in
a hierarchical manner so that the assessment of differ-
ences between the three arch wire materials would take
account of the confounding effect from other variables.
RESULTS
Fifty-six patients requiring upper and lower arch
alignment were included in the study. Data provided
by 51 subjects and 98 arch wires were finally avail-
able, with full records, for analysis. Forty-seven
subjects had usable data for both maxillary and
mandibular arch wires; the remaining four had data
just for one of the arches. Five patients were ex-
cluded from the trial either because of failure of
attendance for arch wire adjustment at the correct
time interval or bracket debond where rebonding
was not performed, as per the protocol, within 48
hours. Table I gives basic study data for the total
American Journal of Orthodontics and Dentofacial Orthopedics Evans, Jones, and Newcombe 35
Volume 114, No. 1

Table II. Reproducibility of 2 and 3 dimensional contact point

measurements

Reproducibility
Number of Mean d2/2 SD
pairs of
Contact point readings 3D 2D 3D 2D

6-5 174 0.001 0.001 0.039 0.036

5-4 177 0.002 0.002 0.052 0.051
4-3 261 0.013 0.005 0.116 0.073
3-2 246 0.009 0.004 0.095 0.063
2-1 264 0.006 0.004 0.076 0.060
1-1 132 0.003 0.002 0.057 0.044
Fig. 2. Mean three dimensional tooth movement for
anterior and posterior segments of dental arches.
Table III. Reproducibility of 2 and 3 dimensional
measurements for upper arch, lower arch and whole sample

Reproducibility
Mean d2/2 SD

Arch Number 3D 2D 3D 2D

Upper 603 0.008 0.003 0.092 0.056

Lower 651 0.005 0.003 0.072 0.060
All 1254 0.006 0.003 0.082 0.058

numbers of each arch wire used and the time span in

Fig. 3. Mean two dimensional tooth movements for
days for the use of each wire during the clinical trial. anterior and posterior segments of dental arches.
Measurement error
Table IV. Mean values for 3 dimensional tooth movements
Reproducibility for two- and three-dimensional
measurements are expressed in Tables II and III. Arch movement (mm 3 100)
Sig of F
Values are expressed for individual tooth point
wire A wire B wire C F (p value)
reproducibility, the mean upper, and mean lower
arch reproducibility, and mean value for all inter- 6-6 441 414 502 0.63 0.54
0-4 wks 3-3 204 232 252 1.70 0.19
tooth distances in both arches, where the reproduc- 3-6 238 182 250 0.18 0.83
ibility standard deviation is given as =Ed2/2n. 6-6 349 290 336 1.93 0.16
All values in three dimensions show acceptable 4-8 wks 3-3 141 129 139 0.45 0.64
3-6 208 161 197 1.33 0.27
unit values for reproducibility. Measurements in the 6-6 612 573 662 0.30 0.74
lower arch are open to less error than the upper, but 0-8 wks 3-3 263 294 298 1.19 0.31
the difference is slight. Figures for specific “digitized 3-6 349 279 365 0.05 0.95

points” on the teeth suggest similar values of mea-

surement error in each case, with measurements
involving the canine contact point recognition giving
the greatest error. For two-dimensional readings
reproducibility is once again acceptable. Overall
values of standard deviation (SD) for the whole
population demonstrate the reduction in error im-
puted when one dimension has been discarded (the
vertical axis alignment). The upper and lower arches
have closer correlation for two-dimensional mea-
surement error. For intertooth measurement repro-
ducibility is once again improved, but measurements
involving the use of canine contact points continue
to yield greater error.
Comparative arch wire performance: Two- and
three-dimensional measurements of change in
tooth alignment
Figs. 2 (three dimensions), and 3 (two dimen-
sions) show the main data for mean units of move-
ment of the whole arch, the anterior arch, and the
posterior arch for each wire type. This information
is depicted in graphic, and tabulated form in Tables
IV and V, for each time period.
Analysis of Variance (ANOVA) values are in-
cluded in a hierarchical manner, taking account of
the main effects listed previously. The conventional
36 Evans, Jones, and Newcombe American Journal of Orthodontics and Dentofacial Orthopedics
July 1998

Table VI. Mean 3 dimensional tooth movement for

upper/lower arches

Arch movement
(mm 3 100)
Sig of F
Upper Lower F (p value)

6-6 514 395 5.15 0.02

0-4 wks 3-3 267 193 6.88 0.01
3-6 247 202 2.11 0.15
6-6 329 322 0.18 0.67
4-8 wks 3-3 148 125 4.94 0.03
3-6 181 198 1.00 0.32
Fig. 4. Mean three dimensional differential dental arch 6-6 674 562 3.43 0.07
0-8 wks 3-3 338 233 10.4 0.02
alignment, comparison of arches.
3-6 337 329 0.15 0.70

Table VII. Mean 2 dimensional tooth movement for

upper/lower arches

Arch movement
(mm 3 100)
Sig of F
Upper Lower F (p value)

0-4 wks 6-6 486 382 6.09 0.02

3-3 234 181 4.33 0.04
3-6 252 201 2.73 0.10
6-6 329 312 0.62 0.43
4-8 wks 3-3 145 121 4.79 0.03
Fig. 5. Mean two dimensional differential dental arch 3-6 184 191 0.50 0.48
alignment, comparison of arches. 6-6 643 537 4.82 0.03
0-8 wks 3-3 297 222 7.06 0.01
3-6 346 315 1.10 0.30
Table V. Mean values for 2 dimensional tooth movements

Arch Movement (mm 3 100)

Sig of F tion is depicted in graphic and tabulated form
wire A wire B wire C F (p value) (Tables VI and VII) for each time period. ANOVA
6-6 463 371 464 0.20 0.82 was used to calculate the level of significance of
0-4 wks 3-3 195 193 231 1.10 0.33 differences in the degree of tooth movement for
3-6 267 177 233 0.28 0.75
upper versus lower arch. The main effects studied
6-6 349 292 319 2.42 0.10
4-8 wks 3-3 138 133 128 0.12 0.88 here were subject, order, and upper or lower arch. It
3-6 211 159 191 2.10 0.13 is important to note here that the arch wire type was
6-6 632 530 605 0.05 0.95
not considered. Therefore, figures for significance
0-8 wks 3-3 254 258 265 0.50 0.58
3-6 377 271 339 0.50 0.58 given by ANOVA relate to differences in the degree
of tooth movement regardless of the wire used.
For three-dimensional tooth alignment mea-
level of significance (p , 0.05), has not been sures, a statistically significant difference between
achieved between arch wires for either two- or arches is noted for whole dental arch tooth move-
three-dimensional measurements, when considering ment during the first 4 weeks and for anterior arch
mean alignment of the whole arch or anterior/ alignment throughout the 8 week trial period. The
posterior segments alone. difference in movement of the anterior segment for
the 0 to 8 week period is highly statistically signifi-
Upper versus lower arch analysis: Two and three cant at p , 0.001. The posterior segment shows no
dimensional measurements significant difference in upper versus lower arch
Figs. 4 (three dimensions) and 5 (two dimen- movement.
sions) show the mean units of movement for the In two dimensions, where the element of vertical
whole arch, the anterior segment, and the posterior tooth alignment is eliminated, a statistically signifi-
segment for the upper or lower arch. The informa- cant difference in horizontal tooth movement is
American Journal of Orthodontics and Dentofacial Orthopedics
Volume 114, No. 1
noted for whole arch alignment over both the 0 to 4
and 0 to 8 week periods, and also for anterior arch
alignment throughout the period of the clinical trial.
The posterior segment alignment shows no signifi-
cant difference in a comparison of upper versus
lower dental arches.
DISCUSSION
Measurement error and the ‘modified point
identification’ technique for the assessment of
tooth alignment
The intention in using bracket line angles and
incisal edges as fiducial points was to limit point
identification measurement error, particularly for
three-dimensional readings, and hence improve
point reproducibility. The use of projected planes
improves point digitization reproducibility, but the
validity of using such a two-dimensional measure-
ment technique to measure alignment and leveling
is questionable because, by definition, the vertical
element of any tooth movement is ignored.
Three-dimensional measurement reproducibility
Whole arch and individual digitized tooth point
reproducibility. The SD of the measurement error
for the whole arch was found to be 0.082 mm, which
is acceptable for the purposes of this study, with
values ranging from 0.039 mm at first molar to
second premolar, to 0.12 mm at the first premolar to
canine intertooth distances. From molar to canine,
the fiducial points were taken as the occlusal line
angles of the brackets. The poorer reproducibility at
the premolar-canine contact point is difficult to
explain but may be related to a greater “reading”
error in the vertical plane at the corner of the arch,
where the canine is situated.
Upper and lower arch reproducibility. The vari-
ables measured for the upper and lower arches also
show acceptable reproducibility. Upper arch mea-
surements were subject to greater error than in the
lower. This is likely to be a result of increased
interbracket span in the upper arch providing
greater contact point distance and therefore a
greater chance of measurement error.
Two-dimensional measurement reproducibility
This data had the vertical element of the tooth
alignment measure removed.
Whole arch and individual contact point reproduc-
ibility. As expected, the SD of measurement error
values are improved when compared with three-
dimensional measurements. The two-dimensional
value is 0.058 mm for the whole arch, compared with
Evans, Jones, and Newcombe 37
0.082 mm for the three-dimensional equivalent. For
individually digitized points, all two-dimensional
values show improved reproducibility over their
three-dimensional counterparts, and the range of
values is also reduced from 0.036 mm at first molar
to second premolar to 0.073 mm at the first premo-
lar to canine intertooth distances.
Although the two-dimensional values represent
an improvement, this is slight, and certainly does not
invalidate the use of three-dimensional measure-
ments to evaluate tooth movement. Because two-
dimensional values ignore the contribution of the
vertical plane one could only expect there to be
greater variation in three dimensions.
Upper and lower arch reproducibility. Once again
reproducibility is improved for two-dimensional
measurements of both arches, but this is more
marked for the upper arch measurements. This
strongly suggests that vertical error plays a greater
role in error contribution for the upper arch.
Comparison with previous arch alignment studies
In a previous similar study24 the measurement
reproducibility SD was 0.167 mm. The value in the
current study of 0.082 mm compares very favorably
and suggests that the occlusal line angle measure-
ment of brackets used to assist point identification
does offer an improvement in reproducibility.
The study by West25 (1992) used the reflex
microscope to a similar regimen and analyzed re-
producibility for whole arch and individual anatomic
contact point measurements in two and three di-
mensions with pooled standard deviations for trip-
licate readings. In three dimensions, the values for
SD for the upper arch averaged 0.392 mm, and for
the lower 0.322 mm. In two dimensions, these values
were 0.216 mm and 0.191 mm, respectively.
Contact point separation can be used to assess
the degree of initial arch crowding, and this allows
analysis of subsequent tooth movement in relation
to the initial crowding. Because the current study
used bracket line angles and incisal edges rather
than anatomic contact points, it was not possible to
determine the amount of tooth movement in rela-
tion to initial crowding; this is one disadvantage of
the technique used in the current study.
Performance trends for wires A, B, and C
ANOVA was used to test for statistically signif-
icant differences in the aligning and leveling capa-
bilities of the three arch wires at the 4 week period
(T4) and the 8 week period (T8). Visual interpreta-
tion of the descriptive statistics for two and three
38 Evans, Jones, and Newcombe
dimensional measurements would seem to suggest
that both the active martensitic wires (wire A and
wire C) were capable of achieving greater arch
alignment than the multistranded steel (wire B) arch
wire, at virtually all time intervals, and particularly
in the posterior aspects of the arch. However,
ANOVA did not demonstrate any clear statistically
significant difference in the aligning capabilities of
wires A, B, and C. Such a finding does not lend
support to manufacturers claims that active marten-
sitic wires are able to bring about a much more rapid
and efficient arch alignment.
It is possible that although the trend demon-
strated is not statistically significant, it may not rule
out a clinically important effect. Certainly, one
would expect the martensitic arch wires to perform
better than multistranded stainless steel arch wires,
as is apparent in the posterior of the dental arches,
but when one visualizes differences in the anterior
segment any such trend is less obvious. In view of
the graded force delivery provided by wire C (Bio-
force), one might also expect this arch wire to show
a trend for greater alignment in the posterior seg-
ment when compared with wire A (medium force/
Titanium Heat Memory Wire), however, this is not
apparent from this data.
Previous authors have also failed to demonstrate
a substantial significant difference in the aligning
capabilities of superelastic nickel titanium and mul-
tistranded stainless steel after 6 weeks of alignment.
As with the current study, the comparisons in these
clinical trials have been between materials, which
show very different properties under bench test
conditions. There is no doubt that active austenitic
and active martensitic arch wires demonstrate ideal
behavior ex vivo. However, the clinical advantages
of the “vastly superior” properties of active marten-
sitic nickel titanium could not be demonstrated in
this extensive clinical trial. Drescher (personal com-
munication) has speculated that in order to properly
use the superelastic properties of NiTi-type wires, it
is necessary to deform them beyond a certain bend-
ing angle when a maligned tooth/bracket is attached
to the wire. This angle can be quite large, as much as
50° to 70°, in order to reach the superelastic plateau
of the wire. It may be that such bending angles are
only rarely encountered in clinical practice and
therefore, most often, one is taking advantage of the
linear elastic, rather than the superelastic properties
of the arch wire. This makes it more difficult in a
random sample of crowding cases to demonstrate
statistically significant differences over a more con-
ventional braided steel arch wire.
American Journal of Orthodontics and Dentofacial Orthopedics
July 1998
Although statistical and clinical significance do
not necessarily equate, it is also possible that the
metabolic response of the periodontal ligament af-
ter force application, and the technique of arch wire
ligation may have varied both between patients and
clinicians. This in turn may have given rise to
significant confounding variables. Although use of a
single operator might in part address one of these
problems, this could be at the cost of introducing a
systematic error that would not be easy to either
identify or quantify. However, any variation in met-
abolic response might be identified, at least in part,
by collecting gingival crevicular fluid before treat-
ment in order to assess base levels of bone turnover
for each individual.31
Variation in alignment of upper and lower arches
Statistical analysis of tooth movement in the
upper arch and lower arch demonstrated that in this
study the mean upper arch movement was generally
greater than for the lower dental arch. This was
statistically significant in comparisons of anterior
segments in both two and three dimensions, al-
though not demonstrated as conclusively for poste-
rior segment readings. This may be related to the
larger interbracket width in the upper arch, increas-
ing the length of arch wire present, giving rise to
increased resilience, and a more physiologic force
distribution. However, one might expect this factor
to be of limited importance for wires demonstrating
hysteresis. Tooth movement may also have been
restricted in the lower labial segment because of the
proximity of cortical bone and the smaller volume of
cancellous bone within.
CONCLUSIONS
1. A controlled, prospective, and randomized clinical
trial was carried out to assess the aligning capabil-
ities of three arch wires commonly used in initial
orthodontic tooth alignment. The three arch wires
investigated in the study were the following: a
multi-stranded stainless steel arch wire (wire B), a
medium force thermodynamic nickel titanium alloy
(wire A), and a graded force delivery thermody-
namic nickel titanium alloy (wire C).
2. The use of an intertooth measurement system
evolved from a previously established method of
assessing tooth alignment was shown to improve
error in both two and three dimensions measure-
ments in comparison with previous studies.
3. ANOVA modeling of the initial tooth movement
produced by three arch wires, taking subject, order,
and upper or lower arch wire as main effects, failed
American Journal of Orthodontics and Dentofacial Orthopedics
Volume 114, No. 1
to show a statistically significant difference in their
aligning capabilities after 4 and 8 week intervals.
4. Regardless of arch wire type, mean tooth move-
ment in the upper labial segment was greater than
the lower, to a statistically significant level. This
may be related to greater interbracket width and
thus the delivery of a more physiologic force in the
upper labial segment.
5. Analysis of the descriptive statistics alone demon-
strated certain trends. Wires A and C (the nickel
titanium alloys) invariably brought about greater
tooth movement at all time intervals when com-
pared with wire B. This might be the expected
outcome taking account of ex vivo properties,
however, this observed effect could not be con-
firmed in the subsequent statistical analysis of the
data.
6. The inability of active martensitic arch wires to
demonstrate their “superior” ex vivo properties at
the clinical level is likely to be related to individual
variations in metabolic response within the peri-
odontal ligament and bone and possibly variations
in ligation technique. It may be that NiTi-type
wires are rarely deformed sufficiently in routine
clinical practice to take advantage of their super-
elastic properties. It is also possible that what
constitutes statistical significance does not equate
with clinical significance and that individual varia-
tion in response has confounded any possible dif-
ferences in alignment related to arch wire type. It
could be beneficial for future studies to assay
gingival crevicular fluid for base levels of bone
turnover before arch alignment and leveling.
REFERENCES
1. Brown IS. The effect of orthodontic therapy on certain types of periodontal defects.
J Perio 1973;44:742-56.
2. Foster TD. A textbook of orthodontics, Oxford, England: Blackwell Scientific
Publications; 1983. p. 181-7.
3. Rock WP, Wilson HJ. Forces exerted by orthodontic aligning arch wires. Br J
Orthod 1988;15:255-9.
4. Storey E, Smith R. Force in orthodontics and its relation to tooth movement. Aust
Dent J 1952;1:11-3.
Evans, Jones, and Newcombe 39
5. Reitan K. Some factors determining the evaluation of forces in orthodontics. Am J
Orthod 1957;43:32-45.
6. Reitan K. Effects of force magnitude and direction of tooth movement on different
alveolar bone types. Angle Orthod 1964;34:244-55.
7. Weinstein S. Minimal forces in tooth movement. Am J Orthod 1967;53:881-903.
8. Ricketts RM, Buch RW, Gugino CF, Hilgers JJ, Schulhof RJ. Bioprogressive
therapy (Book 1). Denver: Rocky Mountain Communications; 1979. p. 93-126.
9. Burstone CJ, Goldberg AJ. Beta titanium: a new orthodontic alloy. Am J Orthod
1980;77:121-32.
10. Reitan K. Biomechanical principles and reaction. In: Graber TH, Swain B, Editors,
Orthodontics: current principles and techniques. St Louis: CV Mosby; 1985. p.
101-92.
11. Burstone CJ. Rationale of the segmented arch. Am J Orthod 1962;48:805-22.
12. Hixon EH, Callow GE, McDonald HW, Tracy RJ. Optimal force, differential force,
and anchorage. Am J Orthod 1969;55:437-57.
13. Caputo AA, Chacanos SJ, Hayashi RK. Photoelastic visualization of orthodontic
forces during canine retraction. Am J Orthod 1974;65:250-9.
14. Hazel RJ, Rohan GJ, West VC. Force relaxation in orthodontic archwires. Am J
Orthod 1984;86:396-402.
15. Andreasen GF, Morrow RE. Laboratory and clinical analysis of Nitinol wire. Am J
Orthod 1978;73:142-61.
16. Goldberg AJ, Burstone CJ. An evaluation of Beta titanium alloys for use in
orthodontic appliances. J Dent Res 1979;58:593-9.
17. Kapila S, Sachdeva R. Mechanical properties and clinical applications of ortho-
dontic wires. Am J Orthod DentoFacial Orthop 1989;96:100-9.
18. Evans TJW, Durning P. Aligning archwires. The shape of things to come? a fourth
and fifth phase of force delivery. Br J Orthod 1996;23:269-75.
19. Andreasen GF. A clinical trial of alignment of teeth using a 0.019“ thermal Nitinol
wire with a transition temperature range between 310C and 450C. Am J Orthod
1980;78:528-32.
20. Andreasen GF, Amborn RM. Aligning, leveling, and torque control: a pilot study.
Angle Orthod 1989;59:51-7.
21. Miura F, Mogi M, and Okamoto Y. New application of superelastic NiTi
rectangular wire. J Clin Orthod 1990;9:544-8.
22. Viazis AD. Clinical applications of superelastic nickel titanium wires. J Clinical
Orthod 1991;6:370-4.
23. Chen R, Zhi YF, Arvystas MG. Advanced Chinese NiTi alloy wire and clinical
observations. Angle Orthod 1992;62:59-65.
24. O’Brien K, Lewis D, Shaw W, Combe E. A clinical trial of aligning archwires. Eur
J Orthod 1990;12:380-4.
25. West AE. A clinical comparison of two initial aligning archwires [MScD thesis].
University of Wales; 1992.
26. West AE, Jones ML, Newcombe RG. Multiflex versus superelastic: a randomized
clinical trial of the tooth alignment ability of initial arch wires. Am J Orthod
Dentofacial Orthop 1995;108:464-71.
27. Evans TJW. Products update. Br J Orthod 1996;23:Suppl.1-4.
28. Jones ML, Staniford H, Chan C. Comparison of superelastic Niti and multi-
stranded stainless steel wires in initial alignment. J Clin Orthod 1990;24:611-3.
29. Jones ML. An investigation into stereophotogrammetric measurement of routine
study casts, and its use in relating palatal cortical adaptation to incisor movement
[MSc Dissertation]. London, England: University of London; 1979.
30. Bhatia SN, Harrison VE. Operational performance of the travelling microscope in
the measurement of study casts. Br J Orthod 1987;14:147-53.
31. Waddington RJ, Embery G, Samuels RH. Characterization of proteoglycan
metabolites in human gingival crevicular fluid during orthodontic tooth movement.
Arch Oral Biol 1994;39:361-8.