A comparison of the rate of space closure using a

nickel-titanium spring and an elastic module: A

clinical study
R. H. A. Samuels, MDSci., M. Orth., S. J. Rudge, M. Orth., and L. H. Mair, PhD
Liverpool, England

A study of the efficiency of space closure after premolar extraction was undertaken, comparing a
nickel-titanium closed coil spring and an elastic retraction module by using sliding mechanics along
an 0.019 x 0.025-inch stainless steel arch wire in 0.022 • O.028-inch preadjusted stainless steel
brackets. The rate of space closure in 17 subjects was analyzed from study models and was found
to be significantly greater and more consistent with the nickel-titanium closed coil springs than with
the elastic modules, in both arches. There were no clinically observable differences in the tooth
positions between the respective techniques. (AM J ORTHOD DENTOFACOFITHOP 1993;103:464-7.)

The particular property of super-elastic
nickel titanium in producing a light continuous force
over a long range of action, compared with previously
available materials, has been well documented, t3 The
possibility that a nickel-titanium closed coil spring, with
a continuous action, might have some a d v a n t a g e s i n
fixed appliance space closure mechanics was investi-
gated and compared with a currently used elastic re-
traction module, providing an intermittent force. 4 How-
ever, concern has previously been expressed that ex-
cessively rapid space closure may also lead to unwanted
effects, such as loss of tooth control and bunching of
soft tissues in the extraction site, with the consequent
reopening of the space later. ~
MATERIALS AND METHOD
Two operators randomly chose 17 patients from a group
of patients being treated with upper and lower fixed appli-
ances. There were 12 girls and 5 boys with a mean age of
14.7 years (range = I 1.I to 17.1 years). All cases had the
four first premolars extracted and were undergoing treatment
with upper and lower fixed appliances, which used pread-
justed brackets with an 0.022 • 0.028-inch slot (Straight
Wire Appliance--Andrews values, A company, San Diego,
Calif.). Nine patients had active space closure measurements
taken from the upper arch, six from the lower arch and two
from both arches. Space closure was carried out on
0.019 x 0.025-inch stainless steel arch wires, with anchor-
age balance between the six anterior teetti against the second
premolar and first molar. Space closure involved both incisor
retraction and molar protraction, to varying degrees, in all
From the University of Liverpool.
Copyright 9 1993 by the American Association of Orthodontists.
0889-5406193151.00 + O.10 8/1/34374
9patients with anchorage balance controlled by anterior torque
adjustments in the arch wire. Extra oral anchorage was not
employed in the treatment of these cases. The arch wires had
been in place for at least I month before the commencement
of space closure, to'allow torque expression to begin. At the
beginningof space closure, closed coil nickel-titaniumsprings
(Sentalloy, GAC International Inc., Central lslip, N.Y.) were
placed in one quadrant (Fig. 1) and an elastic module system'
(Alastik and wire ligature, Unitek Corp., Monrovia, Calif.)
(Fig. 2) in the other quadrant of the same arch. The space
closure mechanics were allocated to the left or right side of
the same arch wire in each patient by random selection, to
achieve within-patient control. The Sentalloy closed coil
springs were medium grade (yellow), stated by the manufac-
turers to provide a force of 150 gm when stretched within a
range of 3 to 15 mm (Fig. 1). The module was activated by
stretching it by 2 to 3 mm, i.e., twice the diameter in every
patient (Fig. 2), providing a starting force of 400 to 450 gm
as measured clinically by a strain gauge. At the subsequent
visit, the force provided by the spring was unchanged,
whereas that of the elastic module had reduced to approxi-
mately zero. The spring or module was attached to the first
molar hook at one end and a crimpable ballhook (American
Orthodontics, Sheboygan, Wis.) on the arch wire just distal
to the canine at the other.
During space closure, patients were recalled for routine
reviews at regular intervals with a mean time interval of 6
weeks (range 5 to 8 weeks), up to a maximum of three visits.
The maximum duration of space closure recording was 161
days. At each recall the elastic modules were replaced,
whereas the springs were left in situ.
Arch wires were checked for any damage, the ends
trimmed to prevent any interference with sliding by the second
molars, and the crimpable ballhooks, which could not initially
be moved with pressure applied by hand instruments, were
checked for slippage.. This was carried out by ensuring that
the ballhook had not moved distal to a mark placed just mesial

464
American Journal of Orthodomics and Dentofacial Orthopedics
Vol.me 103. No. 5
Fig. 1. Closure in one quadrant with nickel-titanium spring attached to crimped hook o n 0.019 x O.025-
inch arch wire.
Fig. 2. Closure in opposite quadrant with elastic module stretched to twice diameter.
to it on the arch wire, made at the time of placement. If there
was a clinically detectable gap between the mark and the
ballhook, the case was excluded. Upper and lower impres-
sions were taken together with clinical photographs. The
impressions were then cast in model stone and stored until
measurement. Space closure was assessed by measuring on
the study models the distance from a point between the central
incisors to a clear reproducible occlusal landmark on the first
molar in each quadrant, with an x-y digitizer interfaced to a
microcomputer. This method of measurement was chosen
because occasionally spaces were found anterior to the canines
and required inclusion in the overall assessment of space
closure. It is appreciated that this method of measurement
will slightly underestimate the true amount of space closure
of both quadrants, but it was thought to be satisfactory for
Samuels, Rudge, and Mair 465
comparison of the two techniques. With repeated measure-
ments, the standard deviation of measurement error was de-
termined to be 0.14 ram. Space closure measurements were
discontinued one visit before the spaces had completely closed
on either side of the arch wire.
STATISTICAL METHODS
As for each patient there were differences in the
interval between successive recall visits, the amount of
space closure with time was analyzed by multiple linear
regression (MLR). 6 This analysis compares the amount
of space closure caused by the two conditions (spring
br-rfiodule). It eliminates the restraint of having fixed
intervals between appointnaents by using all the values
466 Samttels, Rudge, and Mair American Journal of Orthodontics and Dentafacia/ Orthopedics
May 1 9 9 3

MODULE

5.0 ' SPRING I I i t 1~

,ot
; 1/,
A
E 4.0'
EE"
LU
4.0-

rr"

u~ 3.0" U~ 3.0"t
q O
.J
U U 1 1 / 1
w UJ
U
< 2.0' 2.0'~
IX IX 11 / 1 1
; '2"
1.0 9
/Z,,' ' 1.0' 2~
1/"
1 "
/ 1 B=0.038 /2 1 B=0.025
r r2=0.910
40 80 120 160
TIME (DAYS) TIME (DAYS)

Fig. 3. Individual regression lines for spring and module data. 1,2 = Number and position of observed
data points. B = Slope of the regression line. ~ = Coefficient of determination.

Table I. Multiple linear regression statistics for
space closure with springs or modules
vo.o.e i i Be,o i
Time 0.038 1.080
Condition 0.037 0.012
Interaction 0.013 0.310
Arch 0.821 0.033
Dependent variable = Space closure.
B = Partial regression coefficient.
Beta = Beta coefficient.
t = t value of Beta.
p, = significanceof t for each variable.
Condition = Spring or module.
Interaction = Condition x time.
for both conditions throughout the total duration of the
study. The analysis evaluates the partial regression co-
efficients (B~_4) of the independent variables (time, con-
dition, arch) in the following formula:
Space Closure =
B, (time) + B2 (condition) + B~ (interaction) + B, (arch)
where condition = spring or module and interac-
tion = condition • time.
From each B value, and the standard deviation of
the data, the computer calculates the beta coefficients
for each independent variable. Each beta coefficient has
an associated t value from which it is possible to cal-
culate whether the variable had a significant effect in
the overall regression. 6 The interaction was included
because a'significant value for the condition woul/.I-in----"
dicate that there was an equal difference in the amount
of space closure at all times throughout the study. How-
ever, a significant value for the interaction indicates that
the difference increased as time progressed. The test
also evaluates the overall linear relationship of the data
(r 2) to determine if multiple linear regression is a valid
, ,., test for the analysis. In this study the r ~ value was 0.84,
18.33 0.000 which validates the test as a model for the analysis.
0.22 0.828
4.62 0.000 RESULTS
0.82 0.416
The space closure measurements for each recall visit
are shown in Fig. 3 together with the individual regres-
sion lines for the two conditions. These indicate a higher
rate of space closure with the springs. This was con-
firmed by the MLR, which indicated that the beta co-
efficients for time and the interaction were highly sig-
nificant, whereas those for condition and arch were not
significant (Table I). The values for the interaction in-
dicates that the difference in the slopes of the lines
shown in Fig. 3 was highly significant indicating a
greater rate of closure with the springs. The arch was
not a significant variable affecting space closure.
DISCUSSION
This study found a significantly greater rate of clo-
sure of premolar extraction spaces by super-elastic
nickel-titanium closed coil springs compared with an
elastic retraction module in both maxilla and mandible.
In Fig. 3 the results for the springs were more linear
compared with the elastic modules indicating the rate
of space closure was more consistent than with the
modules where it decreased with time. The Sentalloy
springs are manufactured to provide either 100, 150,
or 200 gm when stretched up to 15 mm. The spring
chosen for this study was 150 gm as this has been
suggested as an appropriate force for space closure with
sliding mechanics, a
A low constant force, offered by the nickel-titanium
American Journal of Orthodontics and Dentofacial Orthopedics
Voh~me 103, No. 5
spring may be more biologically acceptable than the
intermittent high force delivery of the elastic modules 9
The latter undergo rapid force degradation in the mouth
of approximately 50% after 4 weeks with consequent
loss of action. 7"8 Consequently, nickel-titanium closed
coil springs may be able to achieve rapid closure by
providing a low constant force, whereas with elastic
modules such an increase in closure rate might involve
increasing the intermittent force or thinning the arch
wires in the buccal segments. Occlusal interferences
can, of course, affect the rate of space closure in the
individual patient as can the balance between molar
protraction and incisor retraction. Although these ef-
fects were not controlled for in the experimental design,
the allocation of device to side by random selection
makes it highly unlikely that these factors could have
caused the higher rate of closure for the springs.
HowevEr, it has been suggested that excessively
rapid space closure might produce adverse effects2
Loss of torque control may result in the upper incisors
being too upright at the end of space closure. Rapid
mesial movement o f the upper molars may allow in-
adequate to~'que control with relative extrusion of theii"
palatal cusps, resulting in functional interferences. Loss
of tip and rotational control adjacent to the extraction
site might produce rolling of the molars and a lateral
open bite, particularly in high angle cases. Furthermore,
excessive gingival hyperplasia in the extraction sites
might prevent full space closure or allow spaces to
reopen after treatment.
A clinical evaluation of tooth position and angula-
tion was undertaken 6 months after the completion of
space closure by examination of study models and or-
thopantomographs taken at the time that closure was
achieved. These were independently examined by two
investigators both unaware as to the side allocation of
spring or module. No clinically detectable difference
in either the tooth position or angulation produced by
the two systems was found. During treatment, the prob-
lems of loss of torque control and excessive gingival
tissue accumulating in the extraction sites were not en-
countered, nor the tendency toward lateral open bite
seen. A visual study of the orthopantomographs at the
end of the treatment showed no clinically significant
difference in the root lengths between the quadrants
tested. Perhaps these unwanted effects were not found
because the continuous force produced by the nickel-
titanium spring is low compared with the higher inter-
mittent force of the module. There was no obvious
difference in the amount of food debris accumulation
between the two systems, and no patient reported any
"--Halton General Hospital
discomfort with the springs. Routine maintenance of
the springs required less chairside time than replace-
ment of the modules.
Samttel$, Rudge, attd Mair 467
The difference in the rates of space closure achieved
by nickel-titanium springs and elastic modules could be
used to create centerline changes during space closure
when placed in the appropriate quadrants. This tech-
nique could therefore be another mechanism for the
correction of dental centerline discrepancies and achiev-
ing asymmetric space closure.
Further investigations to compare the rates of space
closure with nickel-titanium springs of different force
levels, i.e., 100, 150, and 200 gm, would be of interest
so as to explore their uses. An evaluation of different
force levels and their effects on adverse tooth move-
ments would also be advantageous.
CONCLUSIONS
1. The use of super-elastic nickel-titanium coil
springs resulted in a significantly greater and
more consistent rate of space closure than elastic
modules.
. When examined clinically, there was no differ-
ence in tooth position produced by the two sys-
tems after space closure.
3. There was no evidence of greater patient dis-
comfort with the springs.
We gratefully acknowledge the help of Mr. R. Milnes in
the collection of some of the material used in the study.
REFERENCES
19 Miura F, Mogi b,l, Ohura Y, Hamanaka H9 The super-elastic
property of the Japanese NiTi alloy wire for use in orthodontics.
AM J OR'nIOD DENTOFACORTIIOP 1986;90:1-10.
2. Miura F, Mogi M, Ohura Y. The super-elasticJapanese NiTi alloy
wire for use in orthodontics. Part 2. Introduction of the direct
elastic resistance heat treatment method. Eur J Orthod 1988;
10:187-91.
39 Miura F, Mogi M, Ohura Y, Karibe M9The super-elasticJapanese
NiTi alloy wire for use in orthodontics. Part tll. Studies on the
Japanese NiTi alloy coil springs. A~.IJ ORTIIODDENTOFACORTIIOP
1988;94:89-96.
4. McLaughlinRP, Bennett JC. The transition from standard edge-
wise to preadjusted appliancesystems. J Clin Orthod 1989;23:142-
53.
5. Bennett JC, McLaughlin RP. Controlled space closure with a
preadjusted appliance system. J Clin Orthod 1990;24:251-60.
6. Norusis MJ. Introductory statistics guide. Chicago: SPSS~
1983:135-74.
7. Kuster R, Ingervall B, Burgin W. Laboratoryand intra-oral tests
of the degradation of elastic chains. Eur J Orthod 1986;8:202-8.
8. Rock WP, Wilson HJ, Fisher SE. Force reduction of Orthodontic
Elastomeric chains after one month in the mouth. Br J Orthod
1986;13:t47-50.
Reprhzt requests to:
Dr. S. J. Rudge
Orthodontic Department
Runcorn
Cheshire WA7 2DA
England, UK