SYSTEMATIC REVIEW

Effectiveness of orthodontic treatment with

functional appliances on mandibular growth in
the short term
Elvira Marsico,a Elda Gatto,b Maryalba Burrascano,a Giovanni Matarese,c and Giancarlo Cordascod
Messina, Italy

Introduction: The aim of this study was to analyze the current literature for the best evidence (randomized clin-
ical trials) about the efficacy of functional appliances on mandibular growth in the short term. Methods: A survey
of articles published up to September 2009 was performed by using the following electronic databases:
PubMed, Embase, Ovid Medline, Cochrane Central Register of Controlled Trials, Web of Science, LILACS,
and Google Scholar. The reference lists of the retrieved articles were hand-searched for possible missing arti-
cles. No language restriction was applied during the identification of the published studies. A methodologic scor-
ing process was developed to identify which randomized clinical trials were stronger methodologically. The
selection process and the quality assessment were undertaken independently and in duplicate by 2 authors.
A meta-analysis was attempted by using random-effects models. Clinical and statistical heterogeneity was
examined, and a sensitivity analysis was performed. Results: Electronic searches identified the following items:
146 articles were retrieved from PubMed, 45 from Cochrane Central Register of Controlled Trials, 29 from Ovid,
42 from LILACS, 628 from Web of Science, and 1000 from Google Scholar. Thirty-two articles fulfilled the
specific inclusion criteria and were identified as potentially appropriate randomized clinical trials to be
included in this meta-analysis. Only 4 articles, based on data from 338 patients (168 treated vs 170 controls)
with Class II malocclusion in the mixed dentition, were selected for the final analysis. The quality analysis of
these studies showed that the statistical methods were at the medium-high level. The outcome
measurements chosen to evaluate the efficacy of the various functional appliances were Co-Pg, Pg/Olp 1
Co/Olp, and Co-Gn and the values were annualized and standardized to a uniform scale with the
standardized mean differences (SMD). The results of the meta-analysis from the random-effects model
showed a statistically significant difference of 1.79 mm in annual mandibular growth of the treatment group
compared with the control group (SMD 5 0.61, 95% CI, 0.30 to –0.93; chi-square test, 5.34; 3 df; P 5 0.15;
I2 5 43.9%; test for overall effect, Z 5 3.83 and P 5 0.0001). The sensitivity analysis showed a substantially
similar outcome of 1.91 mm (SMD 5 0.65, 95% CI, 0.25 to 1.25; chi-square test, 4.96; 2 df; P 5 0.08; I2 5
59.7%; test for overall effect, Z 5 3.19 and P 5 0.001). Conclusions: The analysis of the effect of treatment
with functional appliances vs an untreated control group showed that skeletal changes were statistically signif-
icant, but unlikely to be clinically significant. (Am J Orthod Dentofacial Orthop 2011;139:24-36)

O
ne of the most controversial topics in orthodon- II malocclusion, mandibular retrusion seems to be a ma-
tics relates to the effectiveness of functional ap- jor contributing factor; it occurs in about one third of
pliances on mandibular growth. In skeletal Class the population.1 Functional appliances encompass
a
a range of removable and fixed devices that are designed
Private practice, Messina, Italy.
b
Research fellow, Department of Orthodontics, School of Dentistry, University of
to alter the position of the mandible, both sagittally and
Messina, Messina, Italy. vertically, to induce supplementary lengthening of the
c
Assistant professor, Department of Orthodontics, School of Dentistry, University mandible by stimulating increased growth at the condy-
of Messina, Messina, Italy.
d
Professor, head, and chairman, Department of Orthodontics, School of Den-
lar cartilage.2-5
tistry, University of Messina, Messina, Italy. Experiments have demonstrated that appliances that
The authors report no commercial, proprietary, or financial interest in the prod- position the mandible anteriorly stimulate significant man-
ucts or companies described in this article.
Reprint requests to: Elvira Marsico, Vle Boccetta, 14, 98122 Messina, Italy;
dibular growth by condyle remodeling in animal models,
e-mail, elvira.marsico@alice.it. but the effects produced in humans are not the same.6-9
Submitted, December 2009; revised and accepted, April 2010. Evidence shows that favorable growth responses are not
0889-5406/$36.00
Copyright Ó 2011 by the American Association of Orthodontists.
always achieved with functional therapy; some authors
doi:10.1016/j.ajodo.2010.04.028 reported increases in overall mandibular length10-14 and
24
Marsico et al 25

changes in the amount of condylar growth,15-18 but others

Table I. PubMed search history
believe that mandibular length cannot be altered by such
therapy.19-21 It has been claimed that most of the Keywords Results
correction of the malocclusion is due to dentoalveolar (1) Randomized controlled trial 335,072
changes with a small but statistically significant amount (2) Randomized controlled trials 338,843
(3) Random allocation 66,497
of skeletal effects.3,22,23
(4) Double blind 123,934
There are also controversies concerning the effects of (5) Double blind method 103,271
functional appliances on the maxilla. Many studies indi- (6) Single blind 29,535
cate that forward growth of the maxilla might be in- (7) Single blind method 24,124
hibited,21,24 but other authors stated that there is no (8) #1 OR #2 OR #3 OR #4 OR #5 OR #6 OR #7 427,734
(9) Class II malocclusion 4387
appreciable effect on the position of the maxilla.3,25-27
(10) Orthodontic functional appliances 1599
All of these conflicting claims could be attributable to (11) #9 OR #10 5386
retrospective study designs or methodologic limitations (12) #8 AND #11 169
such as small samples, inadequate or no control group, (13) #8 AND #11, limit: humans 146
dishomogeneity of the groups for sex and age at the start
of therapy, and different lengths of treatment.28 therapy of Class II malocclusion; (4) had a comparable
In addition, there is a lack of long-term studies on the untreated control group; (5) analyzed treatment effects
effect of functional appliances to evaluate the stability not confounded by additional and concomitant treat-
of skeletal changes because of the great difficulties in re- ments (headgear, extractions, or fixed appliances);
cruiting patients after treatment.12 (6) used cephalometric analysis at the start of treatment
The aim of this study was to analyze results from ran- and just after removal of the functional appliances; and
domized controlled trials (RCTs) in the literature con- (7) concerned with mandibular anteroposterior changes
cerning Class II functional therapy to evaluate the measured by using the anatomic condylion.
efficacy of functional appliances on mandibular growth. Articles were not selected if they did not meet the in-
clusion criteria, if they did not relate to this topic, or if
MATERIAL AND METHODS
they related but had a different aim. Abstracts, labora-
A survey of articles published up to September 2009 tory studies, descriptive studies, individual case reports,
about the effects of functional appliances on mandibular series of cases, reviews, studies of adult patients, con-
growth was performed by using several electronic trolled clinical trials, retrospective longitudinal studies,
databases: PubMed, Ovid (Ovid Medline and Embase), and meta-analyses were excluded. Articles reporting in-
Cochrane Central Register of Controlled Trials, Web of terim outcomes or updates were considered only once
Science, LILACS, and Google Scholar. All electronic because they related to the same samples of subjects.
searches were conducted on September 30, 2009. The RCTs including patients who had received previous or
search strategy for PubMed was conducted according to concomitant treatment for their Class II malocclusion
Cochrane Collaboration guidelines (Table I).29 The key- were also excluded.
words used to identify the corresponding studies in the
other databases were Class II malocclusion and ortho- Data abstraction
dontic functional applianc*. A more selective search strat-
The selection process was done by 2 independent re-
egy for Google Scholar was performed with the following
viewers (E.M. and M.B.) using separate printed forms.
terms: Class II malocclusion, orthodontic functional
The data extracted from each article were compared
appliances, and randomized clinical trial. Electronic
and discussed to resolve any discrepancies and to reach
search results are given in detail in Table II. The reference
unanimous consensus. Cohen’s kappa score was calcu-
lists of the retrieved articles were also hand-searched to
lated to determine interrater reliability.
identify any articles to be included in this evaluation
that might have been missed in the electronic biblio-
graphic databases. No language restriction was applied Validity assessment
during the identification process of published studies. Because high-quality methodologic studies can pro-
vide more reliable conclusions, the current Cochrane
Selection criteria guidelines for assessing quality were adopted to evaluate
To be included in our study, each article had to fulfil the risk of bias of the trials included.30 The quality assess-
the following requirements: (1) related human clinical ment was undertaken independently and in duplicate by
trials; (2) included a randomized selection of the sam- using separate printed forms by the 2 reviewers, and
ples; (3) concerned with functional appliances in the agreement was assessed with the kappa statistic.

American Journal of Orthodontics and Dentofacial Orthopedics January 2011  Vol 139  Issue 1
26 Marsico et al

Table II. Search results from various electronic databases

Not related to the topic or
Database Keywords Results related with different aim No RCTs Selected
PubMed Search history as described by 146 102 16 28
Table I
Cochrane Central Register of (1) Orthodontic functional 45 12 32 1
Controlled Trials appliances; (2) Class II
malocclusion; #1 AND #2;
Ovid (1) Orthodontic functional 29 11 17 1
appliances; (2) Class II
malocclusion; #1 AND #2;
Web of Science (1) Orthodontic functional 628 412 198 18
applianc*; (2) Class II
malocclusion; #1 AND #2;
Google Scholar (1) Orthodontic functional 1000 883 95 22
applianc*; (2) Class II
malocclusion; (3)
randomized clinical trials;
#1 AND #2 AND #3;
LILACS (1) Orthodontic functional 42 10 32 0
applianc*; (2) Class II
malocclusion; #1 AND #2;
Hand search Bibliographies of RCTs, RCTs 1 1
known to the authors
before this study, and RCTs
encountered during
searches for other projects

Study characteristics deviations of the continuous outcomes were used to

All selected RCTs related to orthodontic functional summarize the data for each group.
therapy in children during the transitional dentition The outcome values were standardized to a uniform
stage of development with Class II malocclusion.31-34 scale with standardized mean differences (SMD) and
From the selected articles, the following study were combined by using random-effects models. When
characteristics were recorded: sample size, type of studies use different measurements to evaluate the
functional appliance, sex and age, time of treatment or same outcome, SMD (difference in mean outcome be-
observation, time of daily appliance wear, cephalometric tween groups or standard deviations of outcomes
measurements of outcomes, and follow-up. Clinical among participants) can be used in a meta-analysis for
heterogeneity was assessed by examining the types of combining continuous data.
participants and the interventions for the outcome in The significance of any discrepancies in the esti-
each study. mates of the treatment effects from the RCTs was as-
sessed with the Cochrane test for heterogeneity and
Quantitative data synthesis the I2 statistic. We performed a sensitivity analysis to
examine the effect of the study quality assessment on
The effective increase of total mandibular length was the overall estimates of effect.
the outcome chosen to evaluate the efficacy of the various
functional appliances. Cephalometric reference points
and linear measurements of overall mandibular length RESULTS
recorded from each selected study were condylion- Electronic searches identified the following items:
pogonion (Co-Pg), condylion-gnathion (Co-Gn), and 146 articles from PubMed, 45 from Cochrane Central
pogonion-occlusal line perpendiculare 1 condylion- Register of Controlled Trials, 29 from Ovid, 42 from LI-
occlusal line perpendiculare (Pg/Olp 1 Co/OLp). LACS, 628 from Web of Science, and 1000 from Google
Studies that used articulare were excluded because Scholar. Articles that did not relate to our topic or related
this end-point landmark does not measure the real man- with a different aim, and those that were not RCTs, were
dibular increase.32 We did not consider angular mea- excluded, as shown in Table II.
surements because they do not evaluate effective Of the remaining potentially appropriate RCTs, 38 were
mandibular growth. The mean differences and standard duplicates, so 32 were identified as eligible RCTs to be

January 2011  Vol 139  Issue 1 American Journal of Orthodontics and Dentofacial Orthopedics
Marsico et al 27

Manual search Electronic search

Potentially relevant articles identified and

RCTs retrieved by manual
screened for retrieval (PM=146,
search to be included in the
CCRCT=45, Ov=29, WS=628, GS=1000,
meta-analysis (n=1)
LI=42)

Articles excluded, not relate to the

present topic or relate with different aim
(PM=102, CCRCT=12, Ov=11,
WS=412, GS=883, LI=10)

Articles retrieved for more detailed

evaluation (PM=44, CCRCT=33, Ov=18,
WS=216, GS=117, LI=32)

Articles excluded, not randomized

clinical trials (PM=16, CCRCT=32,
Ov=17, WS=198, GS=95, LI=32)

Potentially appropriate RCTs to be

included in the meta-analysis (PM=28,
CCRCT=1, Ov=1, WS=18, GS=22, LI=0)

RCTs excluded, the same references

retrieved in several databases (n=38)

RCTs to be included in the meta-analysis

(n=32)

RCTs withdrawn as reported by Table III

(n=29)

RCTs with usable information, by

outcome (n=3)

PM=PubMed, CCRCT=Cochrane Central Register of Controlled Trials, Ov=Ovid, WS=Web of Science, GS=Google Scholar, LI=LILACS.

Fig 1. Flow chart.

included in this study (Fig 1).32-63 In detail, 6 studies were Scholar, Web of Science, and Cochrane Central Register
retrieved from PubMed; 4 from Google Scholar; 4 from of Controlled Trials (Table III).
both PubMed and Google Scholar; 4 from PubMed and Twenty-nine of the 32 RCTs were withdrawn accord-
Web of Science; 12 from PubMed, Google Scholar, and ing to the following exclusion reasons: (1) lack of an
Web of Science; 1 from PubMed, Google Scholar, Web untreated control group40,46,47,50,53,56,57,59,63; (2) lack
of Science, and Ovid; and 1 from PubMed, Google of random allocation of the untreated control

American Journal of Orthodontics and Dentofacial Orthopedics January 2011  Vol 139  Issue 1
28 Marsico et al

Table III. Studies excluded from the meta-analysis and reasons of exclusion
Reference Databases Study Reason for exclusion
35 Web of Science, PubMed Webster T et al. Associations between Reported interim outcomes or
changes in selected facial updates32
dimensions and the outcome of
orthodontic treatment. Am J
Orthod Dentofacial Orthop
1996;110:46-53.
36 PubMed Cura N, Saraç M. The effect of Concomitant use of headgear with
treatment with the Bass appliance Bass appliance; cephalometric
on skeletal Class II malocclusions: analysis that did not use condylion
a cephalometric investigation. Eur J
Orthod 1997;19:691-702.
37 PubMed Mao J, Zhao H. The correction of Class Concomitant use of headgear with
II, division 1 malocclusion with bionator appliance
bionator headgear combination
appliance. J Tongji Med Univ
1997;17:254-6.
38 Web of Science, PubMed Tulloch JFC et al. Influences on the Related to the present topic but
outcome of early treatment for different aim
Class II malocclusion. Am J Orthod
Dentofacial Orthop 1997;111:
533-42.
39 Google Scholar, PubMed Keeling SD et al. Anteroposterior Patients received previous treatment
skeletal and dental changes after
early Class II treatment with
bionators and headgear. Am J
Orthod Dentofacial Orthop
1998;113:40-50.
40 Google Scholar, Web of Science, Ghafari J et al. Headgear versus No untreated control group
PubMed function regulator in the early
treatment of Class II division 1
malocclusion: a randomized clinical
trial. Am J Orthod Dentofacial
Orthop 1998;113:51-61.
41 Google Scholar, Web of Science, Tulloch JFC et al. Benefit of early Class Reported interim outcomes or
PubMed II treatment: progress report of updates33
a two-phase randomized clinical
trial. Am J Orthod Dentofacial
Orthop 1998;113:62-72.
42 Google scholar, Web of Science, Illing HM et al. A prospective Lack of control group’s randomization
PubMed evaluation of Bass, bionator and
twin block appliances. Part I—the
hard tissues. Eur J Orthod
1998;20:501-16.
43 Google scholar, Web of Science, Ehmer U et al. An international Lack of control group’s
PubMed, Cochrane comparison of early treatment of randomization: the control sample
angle Class-II/1 cases. Skeletal is the same as in Tulloch et al33
effects of the first phase of
a prospective clinical trial. J Orofac
Orthop 1999;60:392-408.
44 PubMed Wheeler TT et al. Effectiveness of early Reported interim outcomes or
treatment of Class II malocclusion. updates39
Am J Orthod Dentofacial Orthop
2002;121:9-17.
45 Web of Science, PubMed Cevidanes LH et al. Clinical outcomes Cephalometric analysis did not use
of Fr€ankel appliance therapy condylion
assessed with a counterpart
analysis.Am J Orthod Dentofacial
Orthop 2003;123:379-87.

January 2011  Vol 139  Issue 1 American Journal of Orthodontics and Dentofacial Orthopedics
Marsico et al 29

Table III. Continued

Reference Databases Study Reason for exclusion

46 PubMed King GJ et al. Comparison of peer No untreated control group
assessment ratings (PAR) from
1-phase and 2-phase treatment
protocols for Class II malocclusions.
Am J Orthod Dentofacial Orthop
2003;123:489-96.
47 Google Scholar, Web of Science, O’Brien K et al. Effectiveness of No untreated control group
PubMed treatment for Class II malocclusion
with the Herbst or twin-block
appliances: a randomized,
controlled trial. Am J Orthod
Dentofacial Orthop 2003;124:
128-37.
48 Google Scholar, PubMed Araujo AM et al. Adaptive condylar Cephalometric analysis did not use
growth and mandibular condylion
remodelling changes with bionator
therapy—an implant study. Eur J
Orthod 2004;26:515-22.
49 Google Scholar Huang G. Twin-block appliance is Summary trial of O’Brien et al34
effective for the correction of Class
II Division I malocclusion during
mixed dentition. J Evid Based Dent
Pract 2004;4:222-3.
50 PubMed Banks P et al. Incremental versus No untreated control group
maximum bite advancement during
twin-block therapy: a randomized
controlled clinical trial. Am J
Orthod Dentofacial Orthop
2004;126:583-8.
51 Google Scholar, Web of Science, Tulloch JF et al. Outcomes in a Reported interim outcomes or
PubMed 2-phase randomized clinical trial of updates33
early Class II treatment. Am J
Orthod Dentofacial Orthop
2004;125:657-67.
52 Google Scholar Kalha A. Early treatment with the Summary trial of O’ Brien et al34
twin-block appliance is effective in
reducing overjet and severity of
malocclusion. Is the twin-block
orthodontic appliance effective in
early treatment of developing class
II division 1 malocclusion? Evid
Based Dent 2004;5:102-3.
53 Web of Science, PubMed Gill DS et al. Prospective clinical trial No untreated control group
comparing the effects of
conventional Twin-block and
mini-block appliances: part
1. Hard-tissue changes. Am J
Orthod Dentofacial Orthop
2005;127:465-72.
54 Google scholar, Web of Science, Cevidanes LH et al. Assessment of Magnetic resonance imaging study,
PubMed. mandibular growth and response to no cephalometric analysis
orthopedic treatment with
3-dimensional magnetic resonance
images. Am J Orthod Dentofacial
Orthop 2005;128:16-26.

American Journal of Orthodontics and Dentofacial Orthopedics January 2011  Vol 139  Issue 1
30 Marsico et al

Table III. Continued

Reference Databases Study Reason for exclusion

55 Google Scholar, Web of Science, Cevidanes LH et al. Comparison of Magnetic resonance imaging study,
PubMed relative mandibular growth vectors no cephalometric analysis
with high-resolution 3-dimensional
imaging. Am J Orthod Dentofacial
Orthop 2005;128:27-34.
56 Google Scholar, Web of Science, Efstratiadis S et al. Evaluation of Class No untreated control group
PubMed II treatment by cephalometric
regional superpositions versus
conventional measurements. Am J
Orthod Dentofacial Orthop
2005;128:607-18.
57 Google Scholar Liu Ji-hui. A comparative study of No untreated control group
clinical effects of Twin-Block and
strigth wire appliance combined
with cervical headgear and class II
elastic on the treatment of class II,
division I malocclusion. Chinese
Journal of Traditional & Western
Medicine 2006;7:1925-9.
58 PubMed Karacay S et al. Forsus nitinol flat Concomitant use of fixed edgwise
spring and Jasper jumper appliances
corrections of Class II division 1
malocclusions. Angle Orthod.
2006;76:666-72.
59 Google Scholar, PubMed Lee RT et al. A controlled clinical trial No untreated control group
of the effects of the Twin Block and
Dynamax appliances on the hard
and soft tissues. Eur J Orthod
2007;29:272-82.
60 Google scholar, Web of Science, Dolce C et al. Timing of Class II Reported interim outcomes or
PubMed, Ovid Medline treatment: skeletal changes updates39
comparing 1-phase and 2-phase
treatment. Am J Orthod Dentofacial
Orthop 2007;132:481-9; erratum
in Am J Orthod Dentofacial Orthop
2007;132:727.
61 Google Scholar Ren Y. Very few indications justify Summary trial of Tulloch et al51
early treatment for severe Class II
malocclusions. Evid Based Dent
2004;5:100-1.
62 Google scholar, Web of Science, O’Brien K et al. Early treatment for Reported interim outcomes or
PubMed Class II Division 1 malocclusion updates34
with the Twin-block appliance:
a multi-center, randomized,
controlled trial. Am J Orthod
Dentofacial Orthop 2009;135:
573-9.
63 Google Scholar, PubMed Baccetti T et al. Comparison of 2 No untreated control group
comprehensive Class II treatment
protocols including the bonded
Herbst and headgear appliances:
a double-blind study of
consecutively treated patients at
puberty. Am J Orthod Dentofacial
Orthop 2009;135:698.e1-10.

January 2011  Vol 139  Issue 1 American Journal of Orthodontics and Dentofacial Orthopedics
Marsico et al 31

Table IV. Study charateristics of the selected RCTs

Selected Sample and Treatment or Time of daily Methods
references appliances Sex Age (y) Setting observation time appliance wear of meassurement Follow-up
Jakobsson31 17 act NA 8.5 Karolinska 18 mo 11.5 h/d Co-Pg No
Institutet
1967 18 contr 8.5 (Sweden) 18 mo
Nelson 12 HA 7 M, 10-12.9 University of 18 mo Minimum Cephalometric No
et al32 5F Otago of 14 h analysis,
1993 13 FR-2 7 M, Mean, (New Zealand) 18 mo Co-Go, Go-Pg,
6F 11.6 Co-Pg
17 contr 11 M, 18 mo
6F
Tulloch 53 bio 30 M, 9.4 University of 15 mo Not Cephalometric No
et al33 23 F analysis,
1997 61 contr 35 M, 9.4 North Carolina 15 mo declared Co-Gn, hand-wrist
26 F radiographs
O’ Brien 73 TB NA Mean, National Health 15 mo 24 h/d Cephalometric No
et al34 9.7 Service analysis,
2003 74 contr Mean, (United 15 mo excluding Pg/Olp 1 Co/OLp,
9.8 Kingdom) contact stage of maturation
sports and of cervical
swimming spine analysis

Act, Activator; HA, Harvold activator; FR-2, Fr€ankel function regulator; bio, bionator; TB, Twin-block appliance; contr, controls; M, male;
F, female; NA, not available.

group42,43; (3) lack of cephalometric analysis54,55; (4) found in time of daily appliance wear, probably due to
cephalometric analysis that did not use anatomic the different functional appliances tested. Every selected
condylion45,48; (5) simultaneous use of additional RCT analyzed a different type of functional appliance
treatments36,37,39,58; (6) progress reports35,41,44,51,60,62; (activator, bionator, Twin-block, Frankel 2, Harvold acti-
(7) summary trials49,52,61; or (8) related to this topic vator), but they had similar mechanisms of action. Treat-
but with a different aim.38 The reasons for excluding ment or observation times ranged from 15 months33,34 to
these studies are given in detail in Table III. Only 3 arti- 18 months.31,32 No article reported long-term results.
cles met all eligibility criteria and were selected for the Considerable variability in the outcome measure-
final analysis.32-34 One more RCT with available ments (Co-Pg,31,32 Co-Gn,33 and Pg/Olp 1 Co/OLp34)
outcomes, identified by hand-searching, was included.31 was seen because of the many different types of cepha-
lometric analyses adopted by the investigators. The data
Study characteristics of the selected articles are summarized in Table IV.
We collected data from 338 patients (168 treated vs
170 controls) with Class II malocclusion in the mixed Validity assessment
dentition. Two trials included only subjects with severe Because studies with high methodologic quality
Class II malocclusion and without clinically obvious fa- might provide more reliable conclusions,64 a quality
cial asymmetry.33,34 The samples were heterogeneous evaluation of the selected articles was performed accord-
for the numbers and the ages of participants. ing to the current Cochrane guidelines for assessment of
The numbers of treated subjects and controls ranged risk of bias.30
from 17 to 73 and from 17 to 74, respectively; the mean The following quality criteria of each article were ex-
ages of 2 studies was homogeneous at 9.57 years,33,34 amined: sequence generation; allocation concealment;
whereas the others reported mean ages of 8.531 and blinding of participants, personnel, and outcome asses-
11.632 years. Only O’Brien et al34 and Tulloch et al33 de- sors; incomplete outcome data; selective outcome report-
tected an effective level of skeletal maturity through the ing; and other sources of bias. Allocation concealment
evaluation of cervical spine and hand-wrist radiographs, was adequate for all trials, but 2 failed to generate an ad-
respectively. equate allocation sequence; the outcome assessor was
It was not possible to assess clinical sex heterogeneity appropriately blinded in only 1 study, and incomplete
because, in 2 trials, data were not available about the outcome data were adequately addressed in 3 reports.
sexes of the samples.31,34 Strong heterogeneity was All studies apparently seemed free of other problems

American Journal of Orthodontics and Dentofacial Orthopedics January 2011  Vol 139  Issue 1
32 Marsico et al

Table V. Summary assessment of risk of bias

Blinding of
participants, Incomplete Selective
Sequence Allocation personnel outcome outcome Other
generation: concealment: and outcome data: reporting: sources Risk of bias
yes, no, yes, no, assessors: yes, no, yes, no, of bias: yes, (low, unclear,
Articles unclear unclear yes, no, unclear unclear unclear no, unclear high)
Jakobsson31 No Yes No Yes Yes Yes Unclear
Nelson et al32 No Yes No No Yes Yes High
Tulloch et al33 Yes Yes No Yes Yes Yes Low
O’ Brien et al34 Yes Yes Yes Yes Yes Yes Low
Low, Unlikely to seriously alter the results, all criteria met; unclear, some doubt raised about results, at least 1 criterion partially met; high, seriously
weakens the confidence in the result, at least 1 criterion not met.

Fig 2. Forest plot representing the effect of functional appliances with the random-effects model.

that could cause a high risk of bias. Only 2 RCTs were as- meta-analysis showed a statistically significant differ-
sessed with a low risk of bias33,34; 1 was at high risk of ence of 1.79 mm in the annual mandibular growth of
bias,34 and the last 1 was unclear31 (Table V). the treatment groups compared with the control groups
The quality assessment was undertaken indepen- (SMD 5 0.61, 95% CI, 0.30 to –0.93; chi-square test,
dently and in duplicate by using separate printed forms 5.34; 3 df; P 5 0.15; I2 5 43.9%; test for overall effect,
by the reviewers, and agreement was assessed by using Z 5 3.83 and P 5 0.0001) (Fig 2).
the kappa statistic. Substantial agreement was reached To test how robust the results of this meta-analysis
between the reviewers, with a kappa score of 0.75. were, a sensitivity analysis was performed. A new meta-
analysis was carried out without the study with the
Quantitative data synthesis highest risk of bias,32 with a substantially similar out-
come of 1.91 mm (SMD 5 0.65, 95% CI, 0.25-1.25;
The selected RCTs assessed the same outcome (in-
crease in mandibular length) but measured it in a variety chi-square test, 4.96; 2 df; P 5 0.08; I2 5 59.7%; test
for overall effect, Z 5 3.19 and P 5 0.001) as shown
of ways (Co-Pg,31,32 Co-Gn,33 and Pg/Olp 1 Co/OLp34),
in Figure 3.
so it was necessary to standardize the values to a uniform
scale (SMD) to combine them in a meta-analysis.29
DISCUSSION
All the values were annualized to compare different
treatment time. Finally the dimensionless value of In this meta-analysis, we investigated the current lit-
SMD obtained from the random effects model was mul- erature with best evidence (RCTs) about the efficacy of
tiplied by the standard deviation to find the actual dif- functional appliances on mandibular growth. Evidence
ference in mandibular length.30 The results of the is lacking in this field because of the difficulties

January 2011  Vol 139  Issue 1 American Journal of Orthodontics and Dentofacial Orthopedics
Marsico et al 33

Fig 3. Sensitivity analysis: forest plot representing the effect of functional appliances without the
lower-quality study (Nelson et al32).

associated with conducting high-quality studies.65 De- CI, 0.25 to 1.25; chi-square test, 4.96; 2 df; P 5 0.08;
spite an accurate and wide bibliographic search strategy, I2 5 59.7%; test for overall effect, Z 5 3.19 and P 5
we found only 4 eligible trials.31-34 0.001) as shown in Figure 3. This implies that differences
The literature showed great variability in the mea- in mandibular growth observed between groups of pa-
surements of the effective increase of total mandibular tients allocated to different interventions (controls vs
length with functional therapy; different variables and treated) can, apart of random error, be attributed to
reference points were used because of the several types the therapy under investigation. Because the selected
of cephalometric analysis adopted by the investigators. trials were of high to moderate quality (Table V), with
In this meta-analysis, we included only articles that wide clinical heterogeneity (use of different functional
reported the anatomic condylion. Studies that used ar- appliances on children in different phases of mixed den-
ticulare were excluded because its location is determi- tition, and subjects from 4 countries), the results of this
nated by mandibular position: the forward and meta-analysis are most likely generalizable.
downward posture of the condyles after functional ap-
pliance therapy can alter any measurement taken from Implications for practice
this point, and it could be interpreted as an increase in This meta-analysis showed that the treatment resulted
mandibular length.66-68 in a change of skeletal pattern; however, these effectively
Because the selected studies reported 3 types of mea- small increases of the mandibular length, even if statisti-
surements (Co-Pg,31,32 Co-Gn,33 and Pg/Olp 1 Co/ cally significant, appear unlikely to be very clinically
OLp34), it was necessary to standardize the values to a uni- significant. These data seem to support recent reports
form scale (SMD) to combine them in a meta-analysis.29 that 2-phase treatment has no advantages over 1-phase
The results obtained from the random-effects model treatment.51,62,69 However, several benefits must be
analysis showed a statistically significant difference of attributed to the early treatment of Class II malocclusion
1.79 mm annual mandibular growth of the treatment with functional appliances: prevention of trauma to
groups compared with the control groups (SMD 5 maxillary incisors associated with a large overjet,
0.61, 95% CI, 0.30 to –0.93; chi-square test, 5.34; 3 df; interception of the development of dysfunction,
P 5 0.15; I2 5 43.9%; test for overall effect, Z 5 3.83 psychosocial advantages for the child during an
and P 5 0.0001) (Fig 2). The I2 value described a consis- important formative period of life, stable dentoalveolar
tent percentage (43.9%) of total variation across the correction, and improved prognosis and shorter duration
studies that was due to heterogeneity rather than chance. of treatment with fixed appliances.3,67,70,71
This statistical heterogeneity can be explained with
the different dimensions of the samples and with the dif- Implications for research
ferent sizes of treatment effects observed in the selected The RCT results, as suggested by O’Brien,72 “only
trials. The sensitivity analysis showed a substantially show the average effect of an intervention on an average
similar outcome of 1.91 mm (SMD = 0.65, 95% patient with a condition. This is a useful information to

American Journal of Orthodontics and Dentofacial Orthopedics January 2011  Vol 139  Issue 1
34 Marsico et al

most clinicians who are interested in treating the average 8. Charlier JP, Petrovic AG, Stutzmann J. Effects of mandibular hy-
patient.” But the dramatic changes that occasionally perpropulsion on the prechondroblastic zone of rat condyle. Am
J Orthod Dentofacial Orthop 1969;55:71-4.
occur within the samples should not be ignored.
9. Stockly P, Willert H. Tissue reactions in the temporomandibular
In deciding whether these appliances are useful, joint resulting from anterior displacement of the mandible in the
indeed, it is important to note that the blending of Class monkey. Am J Orthod Dentofacial Orthop 1971;60:142-55.
II divergencies might have affected the outcome of the 10. Mills JR. The effect of functional appliances on the skeletal pat-
study. Class II studies should separate patients into at tern. J Orthod 1991;18:267-75.
11. de Almeida MR, Henriques JF, de Almeida RR, Ursi W. Treatment
least the hypodivergent, neutral, and hyperdivergent
effects produced by Frankel appliance in patients with Class II, di-
phenotypes. Obviously, it would be harder to obtain an vision 1 malocclusion. Angle Orthod 2002;72:418-25.
appropriate sample size, but different rotation patterns 12. Perillo L, Johnston LE, Ferro A. Permanence of skeletal changes after
are associated with different phenotypes and require function regulator (FR-2) treatment of patients with retrusive Class II
different treatments.73 malocclusions. Am J Orthod Dentofacial Orthop 1996;109:132-9.
13. de Almeida MR, Henriques JF, Ursi W. Comparative study of the
The results of the literature search showed great
Frankel (FR-2) and bionator appliances in the treatment of Class II
variability in the cephalometric measurements of the ef- malocclusion. Am J Orthod Dentofacial Orthop 2002;121:458-66.
fective increase of total mandibular length; the investi- 14. Toth LR, McNamara JA Jr. Treatment effects produced by the
gators adopted different cephalometric analyses and twin-block appliance and the Fr-2 appliance of Fr€ankel compared
different variables and reference points. It would be use- with an untreated Class II sample. Am J Orthod Dentofacial Orthop
1999;116:597-609.
ful to form a consensus on the types of measurements
15. Baltromejus S, Ruf S, Pancherz H. Effective temporomandibular
used in orthodontic studies, so that direct comparisons joint growth and chin position changes: activator versus Herbst
between trials can be achieved. treatment. A cephalometric roentgenographic study. Eur J Orthod
Finally, RCTs should be carried out and reported 2002;24:627-37.
according to the Consolidated Standards of Reporting 16. Croft RS, Buschang PH, English JD, Meyer R. A cephalometric and
tomographic evaluation of Herbst treatment in the mixed denti-
Trials (CONSORT) guidelines.
tion. Am J Orthod Dentofacial Orthop 1999;116:435-43.
17. Pancherz H, Ruf S, Kohlhas P. Effective condylar growth and chin
CONCLUSIONS position changes in Herbst treatment: a cephalometric roentgeno-
In this systematic review, we analyzed results from graphic long-term study. Am J Orthod Dentofacial Orthop 1998;
114:437-46.
RCTs in the literature concerning Class II functional ther- 18. Ruf S, Baltromejus S, Pancherz H. Effective condylar growth and
apy to evaluate the efficacy of functional appliances on chin position in activator treatment: a cephalometric roentgeno-
mandibular growth in the short term. graphic study. Angle Orthod 2001;71:4-11.
This meta-analysis showed that, when functional ap- 19. Creekmore TD, Radney LJ. Frankel appliance therapy: orthopedic
pliance treatment is provided in early adolescence, there or orthodontic? Am J Orthod Dentofacial Orthop 1983;83:89-108.
20. Giannelly A, Bronson P, Martignoni M, Bernstein L. Mandibular
are small beneficial changes in skeletal patterns, but growth, condylar position and Fr€ankel appliance therapy. Angle
these are probably not very clinically significant. Orthod 1983;53:131-42.
21. Vargervik K, Harvold EP. Response to activator treatment in Class II
We thank Rita Aveni for her assistance with the data- malocclusions. Am J Orthod Dentofacial Orthop 1985;88:242-51.
base search. 22. Robertson NRE. An examination of treatment changes in children
treated with the functional regulator of Frankel. Am J Orthod Den-
REFERENCES tofacial Orthop 1983;83:299-310.
23. Janson GR, Toruno JL, Martins DR, Henriques JF, de Freitas MR.
1. McNamara JA Jr. Components of Class II malocclusions in children Class II treatment effects of the Frankel appliance. Eur J Orthod
8-10 years of age. Angle Orthod 1981;51:177-202. 2003;25:301-9.
2. Johnson LE. Orthodontics: state of the art, essence of the science. 24. Baccetti T, Franchi L, Toth LR, McNamara JA. Treatment timing for
St Louis: C V. Mosby; 1986. p. 88-99. Twin-block therapy. Am J Orthod Dentofacial Orthop 2003;118:
3. Wieslander L, Lagerstrom L. The effect of activator treatment on Class 159-70.
II malocclusions. Am J Orthod Dentofacial Orthop 1979;75:20-6. 25. Basciftci FA, Uysal T, Buyukerkmen A, San Z. The effects of activa-
4. Moss ML, Salentijn L. The primary role of functional matrices in tor treatment on the craniofacial structures of Class II division 1
facial growth. Am J Orthod Dentofacial Orthop 1969;55:566-77. patients. Eur J Orthod 2003;25:87-93.
5. Frankel R. The treatment of Class II, Division 1 malocclusion with 26. McNamara JA, Bookstein FL, Shaughnessy TG. Skeletal and dental
functional correctors. Am J Orthod Dentofacial Orthop 1969;55: changes following functional regulator therapy on Class II
265-75. patients. Am J Orthod Dentofacial Orthop 1985;88:91-111.
6. Woodside DG, Altuna G, Harvold E, Herbert M, Metaxas A. Primate 27. Courtney M, Harkness M, Herbison P. Maxillary and cranial base
experiments in malocclusion and bone induction. Am J Orthod changes during treatment with functional appliances. Am J Orthod
Dentofacial Orthop 1983;83:460-8. Dentofacial Orthop 1996;109:616-24.
7. McNamara JA Jr, Bryan FA. Long term mandibular adaptations to 28. Tulloch JFC, Medland W, Tuncay OC. Methods used to evaluate
protrusive function: an experimental study in Macaca mulatta. Am growth modification in Class II malocclusion. Am J Orthod Dento-
J Orthod Dentofacial Orthop 1987;92:98-108. facial Orthop 1990;98:340-7.

January 2011  Vol 139  Issue 1 American Journal of Orthodontics and Dentofacial Orthopedics
Marsico et al 35

29. Higgins JPT, Green S, editors. Cochrane handbook for systematic 47. O’Brien K, Wright J, Conboy F, Sanjie Y, Mandall N, Chadwick S,
reviews of interventions 4.2.5 (updated May 2005). Available at: et al. Effectiveness of treatment for Class II malocclusion with
http://www.cochrane.org/resources/handbook/hbook.htm. the Herbst or Twin-block appliances: a randomized, controlled
30. Higgins JPT, Green S, editors. Cochrane handbook for systematic trial. Am J Orthod Dentofacial Orthop 2003;124:128-37.
reviews of interventions 5.0.2 (updated September 2009). Avail- 48. Araujo AM, Buschang PH, Melo ACM. Adaptive condylar growth
able at: http://www.mrc-bsu.cam.ac.uk/cochrane/handbook.htm. and mandibular remodelling changes with bionator therapy. An
31. Jakobsson SO. Cephalometric evaluation of treatment effect on implant study. Eur J Orthod 2004;26:515-22.
Class II Division 1 malocclusions. Am J Orthod Dentofacial Orthop 49. Huang G. Twin-block appliance is effective for the correction of
1967;53:446-57. Class II Division I malocclusion during mixed dentition. J Evid
32. Nelson C, Harkness M, Herbison P. Mandibular changes during Based Dent Pract 2004;4:222-3.
functional appliance treatment. Am J Orthod Dentofacial Orthop 50. Banks P, Wright J, O’Brien K. Incremental versus maximum
1993;104:153-61. bite advancement during Twin-block therapy: a randomized
33. Tulloch JFC, Phillips C, Koch G, Proffit WR. The effect of early in- controlled clinical trial. Am J Orthod Dentofacial Orthop
tervention on skeletal pattern in Class II malocclusion: a random- 2004;126:583-8.
ized clinical trial. Am J Orthod Dentofacial Orthop 1997;111: 51. Tulloch JF, Proffit WR, Phillips C. Outcomes in a 2-phase random-
391-400. ized clinical trial of early Class II treatment. Am J Orthod Dentofa-
34. O’Brien K, Wright J, Conboy F, Sanjie Y, Mandall N, Chadwick S, cial Orthop 2004;125:657-67.
et al. Effectiveness of early orthodontic treatment with the 52. Kalha A. Early treatment with the twin-block appliance is effective
Twin-block appliance: a multicenter, randomized, controlled trial. in reducing overjet and severity of malocclusion. Evid Based Dent
Part 1: dental and skeletal effects. Am J Orthod Dentofacial Orthop 2004;5:102-3.
2003;124:234-43. 53. Gill DS, Lee RT. Prospective clinical trial comparing the effects of
35. Webster T, Harkness M, Herbison P. Associations between conventional Twin-block and mini-block appliances: part 1.
changes in selected facial dimensions and the outcome of or- Hard-tissue changes. Am J Orthod Dentofacial Orthop 2005;127:
thodontic treatment. Am J Orthod Dentofacial Orthop 1996; 465-72.
110:46-53. 54. Cevidanes LH, Franco AA, Gerig G, Proffit WR, Slice DE, Enlow DH,
36. Cura N, Saraç M. The effect of treatment with the Bass appliance et al. Assessment of mandibular growth and response to orthope-
on skeletal Class II malocclusions: a cephalometric investigation. dic treatment with 3-dimensional magnetic resonance images. Am
Eur J Orthod 1997;19:691-702. J Orthod Dentofacial Orthop 2005;128:16-26.
37. Mao J, Zhao H. The correction of Class II, division 1 malocclusion 55. Cevidanes LH, Franco AA, Gerig G, Proffit WR, Slice DE, Enlow DH,
with bionator headgear combination appliance. J Tongji Med Univ et al. Comparison of relative mandibular growth vectors with high-
1997;17:254-6. resolution 3-dimensional imaging. Am J Orthod Dentofacial
38. Tulloch JF, Proffit WR, Phillips C. Influences on the outcome of Orthop 2005;128:27-34.
early treatment for Class II malocclusion. Am J Orthod Dentofacial 56. Efstratiadis S, Baumrind S, Shofer F, Jacobsson-Hunt U, Laster L,
Orthop 1997;111:533-42. Ghafari J. Evaluation of Class II treatment by cephalometric re-
39. Keeling SD, Wheeler TT, King GJ, Garvan CW, Cohen DA, gional superpositions versus conventional measurements. Am J
Carbassa S, et al. Anteroposterior skeletal and dental changes after Orthod Dentofacial Orthop 2005;128:607-18.
early Class II treatment with bionators and headgear. Am J Orthod 57. Liu Ji-hui. A comparative study of clinical effects of twin-block
Dentofacial Orthop 1998;113:40-50. and straigth wire appliance combined with cervical headgear and
40. Ghafari J, Shofer FS, Jacobsson-Hunt U, Markowitz DL, Laster LL. Class II elastic on the treatment of Class II, division I malocclusion.
Headgear versus function regulator in the early treatment of Class Chinese Journal of Traditional & Western Medicine 2006;7:
II division 1 malocclusion: a randomized clinical trial. Am J Orthod 1925-9.
Dentofacial Orthop 1998;113:51-61. 58. Karacay S, Akin E, Olmez H, Gurton AU, Sagdic D. Forsus nitinol
41. Tulloch JFC, Phillips C, Proffit WR. Benefit of early Class II treat- flat spring and Jasper jumper corrections of Class II division 1 mal-
ment: progress report of a two-phase randomized clinical trial. occlusions. Angle Orthod 2006;76:666-72.
Am J Orthod Dentofacial Orthop 1998;113:62-72. 59. Lee RT, Kyi CS, Mack GJ. A controlled clinical trial of the effects of
42. Illing HM, Morris DO, Lee RT. A prospective evaluation of Bass, bi- the twin block and Dynamax appliances on the hard and soft tis-
onator and twin block appliances. Part I—the hard tissues. Eur J sues. Eur J Orthod 2007;29:272-82.
Orthod 1998;20:501-16. 60. Dolce C, McGorray SP, Brazeau L, King GJ, Wheeler TT. Timing of
43. Ehmer U, Tulloch CJ, Proffit WR, Phillips C. An international com- Class II treatment: skeletal changes comparing 1-phase and
parison of early treatment of angle Class-II/1 cases. Skeletal effects 2-phase treatment. Am J Orthod Dentofacial Orthop 2007;132:
of the first phase of a prospective clinical trial. J Orofac Orthop 481-9:erratum in: Am J Orthod Dentofacial Orthop 2007;132:727.
1999;60:392-408. 61. Ren Y. Very few indications justify early treatment for severe Class
44. Wheeler TT, McGorray SP, Dolce C, Taylor MG, King GJ. Effective- II malocclusions. Evid Based Dent 2004;5:100-1.
ness of early treatment of Class II malocclusion. Am J Orthod Den- 62. O’Brien K, Wright J, Conboy F, Appelbe P, Davies L,
tofacial Orthop 2002;121:9-17. Connolly I, et al. Early treatment for Class II Division 1 maloc-
45. Cevidanes LH, Franco AA, Scanavini MA, Vigorito JW, Enlow DH, clusion with the Twin-block appliance: a multi-center, random-
Proffit WR. Clinical outcomes of Fr€ankel appliance therapy as- ized, controlled trial. Am J Orthod Dentofacial Orthop 2009;
sessed with a counterpart analysis. Am J Orthod Dentofacial 135:573-9.
Orthop 2003;123:379-87. 63. Baccetti T, Franchi L, Stahl F. Comparison of 2 comprehensive
46. King GJ, McGorray SP, Wheeler TT, Dolce C, Taylor M. Comparison Class II treatment protocols including the bonded Herbst and
of peer assessment ratings (PAR) from 1-phase and 2-phase treat- headgear appliances: a double-blind study of consecutively
ment protocols for Class II malocclusions. Am J Orthod Dentofacial treated patients at puberty. Am J Orthod Dentofacial Orthop
Orthop 2003;123:489-96. 2009;135:698.e1-10.

American Journal of Orthodontics and Dentofacial Orthopedics January 2011  Vol 139  Issue 1
36 Marsico et al

64. Harrison JE. Clinical trials in orthodontics II: assessment of the 69. Harrison JE, O’Brien KD, Worthington HV. Orthodontic treatment
quality of reporting of clinical trials published in three orthodontic for prominent upper front teeth in children. Cochrane Database of
journals between 1989 and 1998. J Orthod 2003;30:309-15. Systematic Reviews 2007, Issue 3. Art. No.: CD003452. DOI: 10.
65. Papadopoulos MA, Gkiaouris I. A critical evaluation of meta-analyses 1002/14651858.CD003452.pub2.
in orthodontics. Am J Orthod Dentofacial Orthop 2007;131:589-99. 70. O’Brien K, Wright J, Conboy F, Chadwick S, Connolly I, Cook P,
66. Chen JY, Will LA, Niederman R. Analysis of efficacy of functional et al. Effectiveness of early orthodontic treatment with the Twin-
appliances on mandibular growth. Am J Orthod Dentofacial block appliance: a multicenter, randomized, controlled trial. Part
Orthop 2002;122:470-6. 2: psychosocial effects. Am J Orthod Dentofacial Orthop 2003;
67. Aelbers CMF, Dermaut LR. Orthopedics in orthodontics: part I, fic- 124:488-95.
tion or reality—a review of the literature. Am J Orthod Dentofacial 71. Antonarakis GS, Kiliaridis S. Short-term anteroposterior treatment
Orthop 1996;110:513-9. effects of functional appliances and extraoral traction on Class II
68. Cozza P, Baccetti T, Franchi L, De Toffol L, McNamara JA. Mandib- malocclusion. Angle Orthod 2007;77:907-14.
ular changes produced by functional appliances in Class II maloc- 72. O’Brien K. Orthodontic growth modification. Br Dent J 2001;191:170.
clusion: a systematic review. Am J Orthod Dentofacial Orthop 73. Standerwick RG. Not all Class II patients alike. Am J Orthod Den-
2006;129:599.e1-12. tofacial Orthop 2009;136:143.

January 2011  Vol 139  Issue 1 American Journal of Orthodontics and Dentofacial Orthopedics