The European Journal of Orthodontics Advance Access published December 1, 2014

European Journal of Orthodontics, 2014, 1–9

doi:10.1093/ejo/cju063

Randomized controlled trial

Five-year postretention outcomes of three

retention methods—a randomized controlled trial
Gudrun Edman Tynelius, Sofia Petrén, Lars Bondemark and Eva Lilja-Karlander
Department of Orthodontics, Faculty of Odontology, Malmö University, Sweden

Correspondence to: Gudrun Edman Tynelius, Harvigslund Stubbarp 112, S-274 91 Skurup, Sweden.
E mail: gudrunedmantynelius@gmail.com

Summary
Objective: Comparison of three different retention strategies 5 years or more postretention.
Design, Setting, and Participants: Randomized, prospective, single-centre controlled trial. Forty-
nine patients (33 girls and 16 boys) were randomly assigned to one of three retention methods
during 2 years by picking a ballot shortly before start of retention treatment. Inclusion criteria were
no previous orthodontics, permanent dentition, normal skeletal sagittal, vertical, and transversal

Downloaded from by guest on June 18, 2015

relationships, Class I dental relationship, space deficiencies, treatment plan with extractions of four
premolars followed by fixed straight-wire appliance. Maxillary and mandibular Little's irregularity
index (LII), intercanine and intermolar width, arch length, and overbite/overjet were recorded in
a blinded manner, altogether 10 measurements on each patient. Significant differences in means
within groups assessed by t-test and between groups by one-way analysis of variance.
Interventions: Retention methods: removable vacuum-formed retainer (VFR) covering the palate
and the maxillary anterior teeth from canine-to-canine and bonded canine-to-canine retainer in the
lower arch (group V-CTC); maxillary VFR combined with stripping of the lower anterior teeth (group
V-S); and prefabricated positioner (group P).
Results: Maxillary mean LII ranged from 1.8 to 2.6 mm, mean intercanine width 33.6–35.3 mm
with a significant difference between groups V-S and P, mean intermolar width 46.8–47.4 mm and
mean arch length 21.8–22.8 mm. Mandibular mean LII ranged from 2.0 to 3.4 mm with a significant
difference between groups V-S and P, mean intercanine width from 25.4 to 26.6 mm, mean
intermolar width from 40.8 to 40.9 mm and mean arch length from 16.9 to 17.3 mm. Mean overbite
ranged from 1.8 to 2.7 mm and mean overjet from 3.7 to 4.1 mm.
Limitations: A single centre study could be less generalizable.
Conclusions: The three retention methods disclosed equally favourable clinical results.
Trial registration: This trial was not registered.
Protocol: The protocol was not published before trial commencement.

Introduction A recent review by the Cochrane group concluded that to date there
is insufficient evidence to single out any particular retention strategy
A major challenge in orthodontic treatment is to inhibit relapse and
as the preferred method (6): it was recommended that future studies
ensure long-term stability of outcome. Growth, initial crowding, and
should include true randomization, reporting of dropouts, adequate
patient compliance have traditionally been regarded as the main deter-
sample size calculation, and a minimum follow-up period of 3 months.
minants of orthodontic treatment stability (1). As long as the maxilla
There are a number of recent studies of the short-term effects
and mandible are still growing, the position of the teeth and thus the
of different retention strategies. Two randomized controlled trials
result of retention treatment may be affected. Most current knowl-
(RCTs) comparing two appliances have shown equal short-term
edge about long-term stability is based on retrospective studies (2–5).

© The Author 2014. Published by Oxford University Press on behalf of the European Orthodontic Society. All rights reserved.
1
For permissions, please email: journals.permissions@oup.com
2 European Journal of Orthodontics, 2014

effects in retaining the outcome of orthodontic treatment of the

maxilla (7,8). Edman Tynelius et al. (9,10) compared the effect of
three different retention strategies on maintaining the outcome of
orthodontic treatment of both jaws and concluded that all three
were equally successful after 1 and 2 years of retention. However,
to our knowledge there is to date no RCT of the capacity of various
retention methods to counteract relapse in the long term.
The aim of this study, in the form of an RCT, was to evaluate and
compare the effects of three different retention strategies on counter-
acting orthodontic treatment relapse at least 5 years postretention.
The null hypothesis tested was that the three retention procedures
achieve equivalent long-term results.

Subjects and methods

Trial design
The study was an RCT performed at a single centre by one expe-
rienced orthodontic team. The Ethics Committee of Lund/Malmö
University, Sweden, approved the protocol and the informed consent
form (LU515-01).

Participants, eligibility, and setting

The study subjects comprised patients referred to an orthodontic
clinic in the National Health Service (NHS), Ystad, Sweden. The NHS
clinic was responsible for treatment of malocclusions of patients in
the southeast County Council of Scania. The NHS in Sweden offers

Downloaded from by guest on June 18, 2015

free dental care up to the age of 20 years, including orthodontic
treatment for patients with severe malocclusion. Between 2001 and
2007, all the study participants had undergone orthodontic treat-
ment by one experienced orthodontist (GET). Each patient and
parent received oral as well as written information. Signed, written
consent was required before inclusion in the trial.
The following inclusion criteria were met:

• No previous experience of orthodontic treatment

• Permanent dentition
• Normal skeletal sagittal, vertical, and transverse relationships
• Class I molar relationship or a maximum of 3 mm anterior or
posterior deviation
• Space deficiencies in both jaws
• Treatment plan involving extraction of four premolars followed
by fixed straight-wire appliances in both jaws (11)

Interventions
The three retention methods investigated were:

• A removable VFR covering the palate and the maxillary ante-

rior teeth from canine-to-canine and a bonded canine-to-canine
retainer in the lower arch (group V-CTC; Figure 1)
• An identical maxillary VFR as in group V-CTC, combined with
stripping of the mandibular anterior teeth (group V-S; Figure 1)
(2,12)
• A prefabricated positioner covering all erupted teeth in the max- Figure 1. Vacuum-formed retainer (VFR) in the maxilla and a canine-to-canine
illa and the mandible (group P; Figure 1) (13) retainer in the mandible, VFR in the maxilla and stripping of the mandibular
incisors and canines and prefabricated positioner.
The VFRs were made of 2 mm Biolon (Dreve Dentamid GmbH,
Unna, Germany) in a Scheu Ministar press (Scheu-Dental GmbH, Mechanical stripping of the mandibular incisors and canines was
Iserlohn, Germany). performed either by hand with single sided medium and fine metal
The canine retainers consisted of 0.7 mm spring hard wire blades (TP Orthodontics, La Porte, Indiana, USA) or with a back-
(Dentaurum noninium, Dentaurum, Ispringen, Germany) bonded to and-forth reciprocating handpiece with Ortho-Strips System (GAC
the mandibular canines with Transbond LC (3M Unitek Orthodontic International, New York, USA). The method of stripping depended on
Products, Monrovia, California, USA). tooth form, non-triangular or triangular respectively, and was carried
G. Edman Tynelius et al. 3

out either at the appointment 5–6 weeks prior to debonding, or at Blinding

debonding. The reason for this was to avoid inflicting damage to the To achieve blinding when measuring dental casts only one dental
papilla. When incisors were non-triangular, there was a greater risk cast at a time was picked out of its box in a cross-sectional man-
of damage to the papilla with the reciprocating handpiece and, thus, ner by an assistant, had its own protocol not showing any previous
single sided blades manually was the option. The aim of stripping was measurements and then a new dental cast from a different patient
to achieve small but distinct enamel flattening of the contact surfaces. box was picked. Two dental casts from the same patients were never
The reduction of any contact point between two teeth was approxi- measured in connection with each other. As the study was prospec-
mately the thickness of the coarse blade of either system, i.e. 0.22 mm tive with continuous patient visits it was inconvenient to anonymize
for hand stripping or 0.34 mm for EVA-stripping. dental casts from the start of the study.
The preformed positioner (Ortho-Tain Positioner, Ortho-Tain
Inc., Toa Alta, Puerto Rico) was a soft plastic appliance covering all Statistical analysis
erupted teeth.
Sample size calculation
All retention appliances were distributed within 1 hour after debond-
The calculated sample size for each group was based on a significance
ing. The patients in groups V-CTC and V-S were instructed to wear the
level of 0.05 and 80 per cent power to detect a clinically meaningful
VFR 22–24 hours per day for 2 days and nights and then at night for
difference of 2.0 mm (SD = 2.0 mm) of the LII. The power analysis
12 months. In group P, the positioner was to be worn for 30 minutes dur-
showed that at least 16 patients would be required in each group.
ing the daytime and during sleep for 12 months. During day-time wear
To compensate for dropouts in the follow-up study, 25 patients were
the patients were instructed to actively chew into their positioners. The
enrolled in each group. Arithmetic means and standard deviations
second year of retention patients wore their retainers every other night.
(SD) on group level at times corresponding to pre-treatment (T0),
end of active treatment and start of retention (T1), end of 2 years of
Outcomes retention treatment (T2), and end of 5 years or more out of retention
Dental casts were measured on four occasions: before orthodontic (T3) were calculated for each variable.
treatment (T0), when the fixed orthodontic appliance was removed and
the retention appliance inserted (T1), when the retention appliance was
Primary and secondary outcomes
removed after 2 years (T2), and 5 years or more out of retention (T3).
Significant differences in means between groups were assessed by
The same examiner (GET) made the following linear measure-
one-way analysis of variance and within groups by t-test to the pre-
ments on dental casts with an electronic digital caliper (Mauser

Downloaded from by guest on June 18, 2015

vious time point using the Statistical Package for Social Sciences (ver-
Digital 6, Winterthur, Switzerland) to a precision of 0.01 mm:
sion 22.0.0.0, IBM SPSS Inc., Chicago, Illinois, USA). P-values less
• LII in the maxilla and the mandible—the summed displacement than 5 per cent (P < 0.05) were considered statistically significant.
of the anatomic contact points of the upper and lower incisors
and canines not including the distal contact points of the canines Error of method
to premolars (14) Twenty randomly selected dental casts were measured by the same
• Intercanine width in the maxilla and the mandible—the distance examiner (GET) on two separate occasions at an interval of 4 weeks.
between the canine cusp tips The error of the method did not exceed 0.45 mm for any of the 10
• Intermolar width in the maxilla and the mandible—the distance measurements (15). No significant mean differences between the two
between the mesiobuccal cusp tips of the first molars series of records were found using paired t-test.
• Arch length in the maxilla and the mandible—the perpendicular
distance from the midpoint of the incisal edges of the central
incisors to a line joining the mesial anatomic contact points of Results
the first molars
Participant flow
• Overbite—the overlap of upper to lower central incisors
A total of 82 patients fulfilled the inclusion criteria. Seven patients (6
• Overjet—the distance parallel to the occlusal plane from the
girls, 1 boy) declined to participate, and thus 75 patients were rand-
incisal edge of the most labial maxillary incisor to the opposing
omized at the start of the trial. Out of these 75 patients, 26 failed to
mandibular central incisor
complete the trial: three declined to participate and 23 left the district
Data on patient ages and treatment times were retrieved from the or were unable to be contacted. Consequently, 49 patients (33 girls, 16
treatment records. boys) attended their 5-year or more appointment (Figure 2). Out of
these 49 patients, 9 were out of retention more than 5 years (Table 1).
Randomization
The generation of randomization sequence was performed in blocks of Baseline data
15 to ensure that equal numbers of patients were allocated to each of the There were no significant intergroup differences with respect to age
three retention groups. A ballot system was used for allocation, in lots of or active treatment time. The mean active orthodontic treatment
15: five ballot sheets labelled ‘maxillary VFR and bonded mandibular time was 1.7 years (SD = 0.4) and the mean age at start of retention
canine-to-canine retainer’, five labelled ‘maxillary VFR and mandibular (T1) was 14.3 years (SD = 1.5). As there were no significant differ-
interproximal enamel reduction (stripping)’, and five with ‘positioner’ ences between boys and girls in any of the study variables, the data
were placed in a basket by an independent person. The patient was then for the genders were pooled for analysis (Tables 2–4).
allocated to one of the three groups by picking a ballot from the basket
shortly before debonding. When the first basket was empty, the second Harms
basket was prepared, and another 15 ballots were extracted successively No harms were reported. In two cases, patients choose to keep their
as patients were recruited to the study. This procedure was altogether CTC in situ after 2 years of retention and consequently they were
repeated five times and concluded in 75 randomized patients. excluded from the long-term study.
4 European Journal of Orthodontics, 2014

Eligible patients with appliance in both jaws n=82

51 girls and 31 boys

Declined to participate n=7 6 girls and 1 boy

Randomized patients n=75

45 girls and 30 boys

T1 Group V-CTC T1 Group V-S T1 Group P

n=25 n=25 n=25
18 girls and 7 boys 14 girls and 11 boys 13 girls and 12 boys

Unable to Unable to Unable to

reach reach reach
1 girl 2 boys 3 boys

T2 Completed two- T2 Completed two-year T2 Completed two-

year retention retention year retention
n=24 n=23 n=22
17 girls and 7 boys 14 girls and 9 boys 13 girls and 9 boys

Downloaded from by guest on June 18, 2015

Refused to Refused to Refused to
participate participate participate
1 girl and 1 boy 2 boys 3 boys
Left the Left the Left the
district/unable to district/unable to district/unable to
reach reach reach
5 girls and 2 boys 4 girls and 2 boys 2 girls and 4 boys

T3 Completed >5 T3 Completed >5 T3 Completed >5

years without years without years without
retention retention retention
n=16 n=17 n=16
12 girls and 4 boys 10 girls and 7 boys 11 girls and 5 boys

Figure 2. The flow diagram of the trial. Vacuum-formed retainer (VFR) in the maxilla and bonded canine-to-canine retainer in the mandible (group V-CTC), VFR
in the maxilla and stripping of the mandibular incisors and canines (group V-S), and positioner covering erupted teeth (group P) at start of retention (T1), after
2 years of retention (T2) and after 2 years or more without retention (T3).

Table 1. Postretention. treatment (T1), the level of mean LII was below 0.5 mm in all three
groups, with no significant intergroup differences. Small changes in
5 years 6 years 7 years 8 years 9 years
mean LII occurred during the retention period (T1–T2). After 5 years
V-CTC 13 1 1 1 or more out of retention (T3), the mean LII was in group V-CTC
V-S 12 4 1 1.8 mm (SD = 1.4), in group V-S 2.6 mm (SD = 1.5) and in group P
P 15 1 2.3 mm (SD = 1.9) with no significant intergroup differences (Table 2).
Totals 40 6 1 1 1 In all three groups, there was an average increase in mean inter-
canine width of 1 and 2 mm during treatment (T0–T1), with a return
Table showing for how many years the 49 patients had been out of reten- to pre-treatment levels in all groups at T3 with 34.4 mm (SD = 1.7)
tion at T3. in group V-CTC, 35.3 mm (SD = 2.1) in group V-S and 33.6 mm
(SD = 1.4) in group P. Small, albeit significant intergroup differences
Maxilla T0–T3 were found between the V-S and P groups after 2 years of retention
At start of orthodontic treatment (T0), the mean LII was 8.6 mm in the (T2) and this difference remained at T3 (Table 2).
V-CTC group, 11.7 mm in the V-S group, and 8.3 mm in the P group. The mean intermolar width and the mean arch length decreased in
No significant differences were found between the groups. After all groups because of extraction of premolars. The mean intermolar
 G. Edman Tynelius et al.

Table 2. Mean pre-treatment measurements (mm) of the maxilla (T0), at the end of treatment/start of the retention treatment (T1), at the end of 2 years of retention treatment (T2), and after
5 or more years postretention (T3) in the three retention groups: removable vacuum-formed retainer (VFR) and bonded mandibular canine-to-canine retainer (V-CTC), VFR and mandibular
anterior stripping (V-S), and positioner (P). LII, Little’s Irregularity Index.

95% confidence interval for 95% confidence interval for 95% confidence interval for
V-CTC (n = 16) mean V-S (n = 17) mean P (n = 16) mean

Standard Standard Standard Analysis of

Maxilla Mean deviation Lower bound Upper bound Mean deviation Lower bound Upper bound Mean deviation Lower bound Upper bound variance

LII T0 8.6 4.9 5.5 11.6 11.7 5.1 8.7 14.8 8.3 4.1 5.7 11.0 NS
LII T1 0.4*** 0.4 0.2 0.7 0.5*** 0.5 0.3 0.8 0.5*** 0.3 0.3 0.6 NS
LII T2 0.9* 0.9 0.5 1.4 1.4** 1.0 0.9 1.9 1.7** 1.5 0.9 2.5 NS
LII T3 1.8** 1.4 1.1 2.5 2.6** 1.5 1.8 3.3 2.3* 1.9 1.4 3.3 NS
Intercanine width T0 34.5 2.3 33.0 36.0 34.8 3.8 32.6 36.9 35.0 2.7 33.3 36.7 NS
Intercanine width T1 36.3* 1.6 35.4 37.1 36.6* 2.7 35.2 38.0 36.4* 1.7 35.5 37.3 NS
Intercanine width T2 35.1*** 1.6 34.2 35.9 35.9** 2.4 34.6 37.1 34.1*** 1.5 33.3 34.9 0.036* V-S/P
Intercanine width T3 34.4* 1.7 33.5 35.3 35.3* 2.1 34.2 36.4 33.6* 1.4 32.8 34.3 0.025* V-S/P
Intermolar width T0 48.7 2.9 47.2 50.3 49.8 4.6 47.3 52.3 49.8 3.2 48.0 51.6 NS
Intermolar width T1 48.7 1.5 47.9 49.5 48.9 2.9 47.4 50.4 49.7 2.2 48.5 50.8 NS
Intermolar width T2 47.6** 2.0 46.5 48.6 47.7** 3.5 46.0 49.5 48.1*** 2.0 47.0 49.1 NS
Intermolar width T3 46.8** 1.9 45.7 47.8 47.2** 3.5 45.4 48.9 47.4** 1.8 46.4 48.3 NS
Arch length T0 28.8 3.2 27.1 30.5 29.1 3.1 27.4 30.7 29.1 3.1 27.4 31.1 NS
Arch length T1 23.1*** 1.4 22.4 23.9 22.6*** 2.1 21.5 23.6 22.5*** 3.2 20.8 24.2 NS
Arch length T2 23.2 1.7 22.3 24.1 22.5 1.8 21.6 23.4 22.5 1.7 21.6 23.5 NS
Arch length T3 22.8* 1.7 21.9 23.7 22.3 2.2 21.1 23.4 21.8** 1.7 20.9 22.7 NS

Within groups the t-test shows the significance corresponding to the previous time point. NS, not significant.
*P < 0.05, **P < 0.01, ***P < 0.001.
5

Downloaded from by guest on June 18, 2015

 6

Table 3. Mean pre-treatment measurements (mm) for the mandible (T0), at the end of treatment/start of the retention treatment (T1), at the end of 2 years of retention treatment (T2), and after
5 or more years postretention (T3) in the three retention groups: removable vacuum-formed retainer (VFR) and bonded mandibular canine-to-canine retainer (V-CTC), VFR and mandibular
anterior stripping (V-S), and positioner (P). LII, Little’s Irregularity Index.

95% confidence interval for 95% confidence interval for 95% confidence interval for
V-CTC (n = 16) mean V-S (n = 17) mean P (n = 16) mean

Standard Standard Standard Analysis of

Mandible Mean deviation Lower bound Upper bound Mean deviation Lower bound Upper bound Mean deviation Lower bound Upper bound variance

LII T0 6.2 4.0 4.1 8.4 7.0 4.3 4.7 9.3 5.6 3.4 3.8 7.4 NS
LII T1 0.4*** 0.3 0.2 0.6 0.3*** 0.5 0.1 0.5 0.3*** 0.4 0.1 0.6 NS
LII T2 0.9* 0.7 0.5 1.2 1.2** 1.1 0.7 1.8 2.1*** 1.7 1.3 3.0 0.015* V-
CTC/P
LII T3 2.1*** 0.9 1.6 2.5 2.0** 1.9 1.1 3.0 3.4** 2.1 2.3 4.5 0.037* V-S/P
Intercanine width T0 26.2 2.5 24.9 27.6 26.3 2.4 25.1 27.6 26.3 1.8 25.4 27.3 NS
Intercanine width T1 27.3 1.6 26.4 28.1 27.7** 1.9 26.7 28.7 27.3* 1.0 26.8 27.8 NS
Intercanine width T2 27.5 1.6 26.6 28.3 26.8** 2.2 25.7 28.0 26.0*** 1.2 25.3 26.6 NS
Intercanine width T3 26.6** 1.4 25.9 27.4 26.5* 2.2 25.3 27.6 25.4** 1.5 24.6 26.1 NS
Intermolar width T0 44.2 2.1 43.1 45.3 44.8 3.4 43.1 46.6 44.7 2.6 43.4 46.1 NS
Intermolar width T1 41.8*** 1.9 40.7 42.9 42.2** 2.6 40.9 43.5 43.0* 2.1 41.9 44.1 NS
Intermolar width T2 41.9 1.7 41.0 42.8 41.6 2.7 40.2 43.0 41.5* 2.1 40.4 42.6 NS
Intermolar width T3 40.9** 1.7 40.0 41.8 40.8*** 3.1 39.3 42.4 40.9** 2.7 39.4 42.3 NS
Arch length T0 23.7 2.3 22.5 25.0 23.5 2.5 22.3 24.8 23.4 2.5 22.0 24.7 NS
Arch length T1 17.9*** 1.3 17.2 18.6 17.9*** 1.7 17.0 18.8 17.4*** 1.3 16.7 18.1 NS
Arch length T2 18.2 1.9 17.2 19.2 17.7 1.6 16.9 18.5 17.6 1.4 16.9 18.4 NS
Arch length T3 17.1*** 1.7 16.2 18.0 17.3** 1.7 16.4 18.2 16.9*** 1.2 16.3 17.5 NS

Within groups the t-test shows the significance corresponding to the previous time point. NS, not significant.
*P < 0.05, **P < 0.01, ***P < 0.001.
European Journal of Orthodontics, 2014

Downloaded from by guest on June 18, 2015

G. Edman Tynelius et al. 7

width remained stable during the retention period as well during the

Table 4. Mean overbite and overjet (mm) pre-treatment (T0), at the end of treatment/start of the retention period (T1), at the end of the 2 year retention period (T2), and 5 or more years pos-
tretention (T3) in the three retention groups: removable vacuum-formed retainer (VFR) and bonded mandibular canine-to-canine retainer (V-CTC), VFR and mandibular anterior stripping (V-S),
5-year post retention period. It was at T3 in group V-CTC 46.8 mm

Analysis of

0.040* V-
CTC/V-S
(SD = 1.9), in group V-S 47.2 mm (SD = 3.5), and in group P 47.4 mm

variance
(SD = 1.8). No significant intergroup differences were found at T3

NS
NS
NS
NS
NS
NS
NS
(Table 2).
The mean arch length remained stable during the retention

Upper bound
95% confidence interval for
period as well during the 5-year post retention period and was at T3
for group V-CTC 22.8 mm (SD = 1.7), for group V-S 22.3 (SD = 2.2),
and for group P 21.8 mm (1.7) (Table 2).

4.3
3.1

7.0
3.3
4.1
4.2
2.8
3.0
Lower bound
Mandible T0–T3
At start of orthodontic treatment (T0) the mean LII was 6.2 mm in the

mean
V-CTC group, 7.0 mm in the V-S group, and 5.6 mm in the P group:

3.1
2.3

4.4
2.6
3.0
2.4
1.9
1.8
the intergroup differences were not significant. After treatment (T1),
the mean LII was less than 0.4 mm in all three groups, with no sig-

Standard deviation
nificant intergroup differences. Small changes in mean LII occurred
during the retention period (T1–T2). In the P-group a small but sig-
nificantly higher mean LII was apparent compared to the V-CTC,
mean 2.1 versus 0.9 mm. After 5 years or more out of retention (T3),

1.2
2.6
0.7
1.0
1.2
0.9
1.7
0.8
the mean LII in group V-CTC was 2.1 mm (SD = 0.9), in group V-S

P (n = 16)
2.0 mm (SD = 1.9), and in group P 3.4 mm (2.1). The average value

2.9***
Mean
was significantly higher in the P group than in the V-S group (Table 3).

3.5*
2.3*

3.7
5.7
2.4
2.7
3.3
In all three groups, mean intercanine width increased on average
1 mm during treatment (T0–T1) and had returned to pre-treatment

Upper bound
95% confidence interval for
levels at T3 with 26.6 mm (SD = 1.4) in group V-CTC, 26.5 mm
(SD = 2.2) in group V-S and 25.4 mm (SD = 1.5) in group P. No

Downloaded from by guest on June 18, 2015

significant intergroup differences were found after 5 years or more

4.4

4.7
6.9
3.8
3.0
3.3
3.7
2.2
(Table 3).

Lower bound
As was found in the maxilla, mandibular mean intermolar width
and mean arch length decreased during treatment (T0–T1) in all
groups because of extraction of premolars. The mean intermolar

Within groups the t-test shows the significance corresponding to the previous time point. NS, not significant.
mean

3.5
1.9

4.3
2.6
3.3
1.6
1.5
1.9
width was at T3 for group V-CTC 40.9 mm (SD = 1.7), for group
V-S 40.8 mm (SD = 3.1), and for group P 40.9 mm (SD = 2.7). No
Standard deviation

significant intergroup differences were found after 5 years or more

(Table 3).
The mean arch length was at T3 for group V-CTC 17.1 mm
(SD = 1.7), for group V-S 17.3 mm (SD = 1.7), and for group P
16.9 mm (SD = 1.2). No significant intergroup differences were
V-S (n = 17)

1.2
1.3

2.6
1.1
1.1
2.0
0.7
1.0

found after 5 years or more (Table 3). 3.2**

2.5**
Mean

3.8*

4.1
2.6

5.6
2.6
1.8

Overbite and overjet T0–T3

During the observation period, including treatment, the mean over-
Upper bound
95% confidence interval for

bite remained fairly unaltered and was in group V-CTC 1.8 mm

(SD = 1.6), in group V-S 2.6 mm (SD = 1.3), and in group P 2.7 mm
(SD = 0.9). After 5 years or more, no intergroup differences were
4.4
5.6
3.2
3.4
2.2
2.7
2.9
2.6

found (Table 4).

Lower bound

During treatment (T0–T1), the mean overjet decreased signifi-

cantly in all three groups and without any group differences. At the
end of the retention period (T2), the mean overjet was significantly
mean

3.0
3.4
2.6
2.3
1.2
1.6
1.0
1.0

higher in the V-S than in the V-CTC group (3.8 versus 2.9 mm). After
*P < 0.05, **P < 0.01, ***P < 0.001.

5 years or more postretention (T3), there were no significant inter-

Standard deviation

group differences in mean overjet with 3.7 mm (SD = 1.3) in group

V-CTC, with 4.1 mm (SD = 1.2) in group V-S, and with 3.7 mm
(SD = 1.2) in group P (Table 4).
V-CTC (n = 16)
and positioner (P).

1.3
1.6

2.0
0.6
1.1
1.6
0.9
1.1

Discussion
3.7**
Mean

After 5 years or more out of retention, all three retention methods

T1 2.9*
T2 2.9
T0 4.5
T2 1.6
T3 1.8
T0 2.0
T1 2.1

provided good long-term clinical stability. Thus, the null hypothesis

Overbite

Overjet

could not be rejected for the patient groups corresponding to the

T3

inclusion criteria. Consequently, any of the three retention methods

8 European Journal of Orthodontics, 2014

can be recommended in patients with space deficiency in both jaws, known and unknown determinants of outcome are evenly distrib-
when treatment involves extraction of four premolars followed by uted among the subjects. The prospective design also ensures that the
fixed straight-wire appliances. baseline characteristics, treatment progression, and side-effects can
be strictly controlled and accurately observed.
Main findings The standardized treatment of all patients by the same orthodontic
There are a number of studies evaluating stability after orthodontic team makes it a valid comparison of the actual treatment methods.
treatment (3,6,16–19). However, to our knowledge, this is the first In most cases there was substantial pre-treatment crowding (T0). In
RCT specifically designed to evaluate and compare three different all groups, the mean LII after treatment (T1) was below or equal to
retention strategies for counteracting relapse at least 5 years out of 0.5 mm in both jaws, indicating successfully treated cases (Tables 2
retention. Thus, no comparison can be made with previous studies. and 3). However, the mean intercanine width was generally somewhat
Whilst not directly comparable, a few RCTs of short-term stabil- expanded after treatment and relapse could be expected: the mean
ity of orthodontic treatment results have previously been published intercanine widths had returned to pre-treatment values in all groups.
(7,8,20,21). These also demonstrated and confirmed our results that
small but not clinically significant movements of teeth occurred. Limitations
From the second day after debonding, the patients in the V-CTC In a long-term study, the effect of attrition on outcomes must be con-
and V-S groups wore their VFRs only at night. This regimen was sidered and some attrition is inevitable. Hence, in the present study the
based on a report of equal stability of treatment result following attrition rate was acceptable and according to the sample size calcula-
full- or part-time wear of Essix retainers (8). Vacuum-formed Essix tion, the number of subjects remaining in each group was adequate to
retainers have been shown to be superior to Hawley retainers in ensure that the outcomes were not biased by loss of data.
retaining the maxillary anterior teeth and were thus the option of The standardized treatment of all patients by the same ortho-
choice (7). LII increased in patients with canine-to-canine retainers dontic team could be a drawback as the same possible mistakes may
in the present study: this is in accordance with other studies (19). have been made on all patients.
In group V-CTC, it was obvious that the canine-to-canine retainers The use of LII for measuring relapse of tooth positions may
held the intercanine width well at T2, but the extraction sites opened have some limitations: it tends to exaggerate cases with consid-
up during retention treatment, i.e. the arch length increased and the erable irregularity but little length shortage, i.e. a rotated tooth
overjet decreased compared to group V-S. This is probably due to without crowding (14). Another drawback with the LII is that
a proclination of the lower front teeth but it cannot be confirmed it does not include the distal contact point of the canine. This

Downloaded from by guest on June 18, 2015

cephalometrically as we did not take cephalograms at T2. At T3, means that canines can be in a very crowded position without
teeth were free to move without any influence of retention devices LII showing it.
and, thus, the groups grew similar. In any retention treatment with removable appliances, the
The finding that stripping of the mandibular anterior teeth with- responsibility and result lies with the patient and is out of control
out any adjunctive methods was sufficient to retain the treatment to the orthodontist.
result on a short-term basis has not previously been demonstrated At the final registration, all patients in this trial had reached
in an RCT. A retrospective study reported that stripping of lower adulthood. This means that overall growth had ceased but this does
anterior teeth in combination with overcorrection of rotated teeth not imply that changes in occlusion, arch dimensions, and tooth
was sufficient to prevent relapse over a 3-year period (18). Stripping position cannot occur. Thus, the increase in Little's index and dimin-
of lower anterior teeth after debonding instead of wearing fixed or ished intercanine widths found in all groups during the postretention
removable retention appliances would probably have advantages, period could coincide with the normal physiological changes during
not only for the patient but also for the clinician, in terms of cost, the same period of time (22,23).
chair-time, and the issue of lost appliances. New studies are needed to confirm the results of this study.
The positioner is an eruption guidance appliance used in the
early mixed dentition (13). No research has been presented about
the effectiveness of prefabricated positioners for retention of ortho- Conclusions
dontic treatment results. This study showed that the appliance could After 5 years or more out of retention, the three retention meth-
be used in the permanent dentition as a retention device but in the ods had achieved equally favourable clinical results. Thus a maxil-
long-term perspective the positioner held the maxillary intercanine lary VFR combined with a bonded canine-to-canine retainer in the
distance less well than the VFR and likewise the mandibular LII mandible; a maxillary VFR combined with stripping of the man-
compared to the V-S group. Because the positioner is prefabricated, dibular anterior teeth and a prefabricated positioner can all be rec-
it may be a less costly alternative to appliances made by dental tech- ommended. Hence, the clinician is not limited to routine use of a
nicians. However, a drawback is that the fitting cannot be as precise bonded mandibular canine-to-canine retainer: choice of retention
as retention appliances made on individual dental casts. A further method can be individualized, taking into account such variables as
disadvantage is that compliance is essential, but as the subjects were orthodontic diagnosis, the expected level of patient compliance and
randomly allocated, this uncontrolled factor was evenly distributed the patient’s wishes and financial considerations.
among the groups.

Generalizability Funding
An RCT was selected in order to reduce the risk of error from such Scania County Council, Sweden; the Swedish Dental Society and the
factors as selection bias, the clinician's preferred treatment method Faculty of Odontology, Malmö University, Sweden.
and patient compliance. Furthermore, random allocation of sub-
jects reduces bias and confounding variables by ensuring that both Conflict of interest statement. None declared.
G. Edman Tynelius et al. 9

References 12. Joseph, V.P., Rossouw, P.E. and Basson, N.J. (1992) Orthodontic micro-
abrasive reproximation. American Journal of Orthodontics and Dentofa-
1. Melrose, C. and Millett, D. (1998) Toward a perspective on orthodontic
cial Orthopedics, 102, 351–359.
retention? American Journal of Orthodontics and Dentofacial Orthope-
13. Keski-Nisula, K., Hernesniemi, R., Heiskanen, M., Keski-Nisula, L. and Varrela, J.
dics, 113, 507–514.
(2008) Orthodontic intervention in the early mixed dentition: a prospective, con-
2. Boese, L.R. (1980) Fiberotomy and reproximation without lower reten-
trolled study on the effects of the eruption guidance appliance. American Journal
tion 9 years in retrospect: part II. The Angle Orthodontist, 50, 169–178.
of Orthodontics and Dentofacial Orthopedics, 133, 254–260; quiz 328.e2.
3. Little, R.M. (1990) Stability and relapse of dental arch alignment. British
14. Little, R.M. (1975) The irregularity index: a quantitative score of mandib-
Journal of Orthodontics, 17, 235–241.
ular anterior alignment. American Journal of Orthodontics, 68, 554–563.
4. Zachrisson, B.U. (1997) Important aspects of long-term stability. Journal
15. Dahlberg, G. (1940) Statistical Methods for Medical Biological Students.
of Clinical Orthodontics, 31, 562–583.
Allen and Unwin, London, pp. 122–132.
5. Myser, S.A., Campel, P.M., Boley, J. and Buschang, P.H. (2013) Long-term
16. Reitan, K. (1967) Clinical and histologic observations on tooth movement
stability: postretention changes of the mandibular anterior teeth. Ameri-
during and after orthodontic treatment. American Journal of Orthodon-
can Journal of Orthodontics and Dentofacial Orthopedics, 44, 420–429.
tics, 53, 721–745.
6. Littlewood, S.J., Millett, D.T., Doubleday, B., Bearn, D.R. and Worthing-
17. Taner, T., Haydar, B. and Kavuklu, I. (2000) Short-term effects of fiber-
ton, H.V. (2006) Retention procedures for stabilising tooth position after
otomy on relapse of anterior crowding. American Journal of Orthodontics
treatment with orthodontic braces. The Cochrane Library, 3, CD002283.
and Dentofacial Orthopedics, 118, 617–623.
7. Rowland, H., Hichens, L., Williams, A., Hills, P., Killingback, N., Ewings,
18. Aasen, T.O. and Espeland, L. (2005) An approach to maintain orthodontic
P., Clark, S., Ireland, A.J. and Sandy, J.R. (2007) The effectiveness of Haw-
alignment of lower incisors without the use of retainers. European Journal
ley and vacuum-formed retainers: a single-center randomized controlled
of Orthodontics, 27, 209–214.
trial. American Journal of Orthodontics and Dentofacial Orthopedics,
19. Atack, N., Harradine, N., Sandy, J.R. and Ireland, A.J. (2007) Which way
132, 730–737.
forward? Fixed or removable lower retainers. The Angle Orthodontist, 77,
8. Gill, D.S., Naini, F.B., Jones, A. and Tredwin, C.J. (2007) Part-time ver-
954–959.
sus full-time retainer wear following fixed appliance therapy: a rand-
20. Jäderberg, S., Feldmann, I. and Engström, C. (2012) Removable thermo-
omized prospective controlled trial. World Journal of Orthodontics, 8,
plastic appliances as orthodontic retainers—a prospective study of differ-
300–306.
ent wear regimens. European Journal of Orthodontics, 34, 475–479.
9. Edman Tynelius, G., Bondemark, L. and Lilja-Karlander, E. (2010) Evalu-
21. Lindauer, S.J. and Shoff, R.C. (1998) Comparison of Essix and Hawley
ation of orthodontic treatment after 1 year of retention – a randomized
retainers. Journal of Clinical Orthodontics, 32, 95–97.
controlled trial. European Journal of Orthodontics, 32, 542–547.
22. Thilander, B. (2009) Dentoalveolar development in subjects with normal
10. Edman Tynelius, G., Bondemark, L. and Lilja-Karlander, E. (2013) A

Downloaded from by guest on June 18, 2015

occlusion. A longitudinal study between the ages of 5 and 31 years. Euro-
randomized controlled trial of three orthodontic retention methods in
pean Journal of Orthodontics, 31, 109–120.
Class I four premolar extraction cases – stability after 2 years in retention.
23. Tsiopas, N., Nilner, M., Bondemark, L. and Bjerklin, K. (2013) A 40 years
Orthodontics & Craniofacial Research, 16, 105–115.
follow-up of dental arch dimensions and incisor irregularity in adults.
11. McLaughlin, R.P., Bennett, J. and Trevisi, H. (2001) Systemized Orthodon-
European Journal of Orthodontics, 35, 230–235.
tic Treatment Mechanics. Mosby International Ltd, London.