European Journal of Orthodontics, 2024, 46, 1–9

https://doi.org/10.1093/ejo/cjad053
Systematic Review

Vertical effects of cervical headgear in growing patients

with Class II malocclusion: a systematic review and
meta-analysis
Umar Hussain1, , Ahsan Memood Shah2, , Fazli Rabi3, , Alessandra Campobasso4, ,
Spyridon N. Papageorgiou5,*,

Downloaded from https://academic.oup.com/ejo/article/46/1/cjad053/7326801 by guest on 06 May 2024

1
Department of Orthodontics, Saidu College of Dentistry, Khyber Pakhtunkhwa, Swat, Pakistan
2
Department Orthodontics, Khyber College of Dentistry, Peshawar, Khyber Pakhtunkhwa, Pakistan
3
Department of Orthodontics, Saidu College of Dentistry, Swat, Khyber Pakhtunkhwa, Pakistan
4
Department of Clinical and Experimental Medicine, University of Foggia, Via Rovelli 50, 71122 Foggia, Italy
5
Clinic of Orthodontics and Pediatric Dentistry, Center of Dental Medicine, University of Zurich, Plattenstrasse 11, 8032 Zurich, Switzerland
*
Corresponding author. Clinic of Orthodontics and Pediatric Dentistry, Center of Dental Medicine, University of Zurich, Plattenstrasse 11, 8032 Zurich,
Switzerland. E-mail: snpapage@gmail.com

Abstract
Background: Cervical headgear (cHG) has been shown to be effective in Class II correction both with dental and orthopaedic effects but has
traditionally been associated with vertical adverse effects in terms of posterior mandibular rotation.
Objective: To assess the treatment effects of cHG treatment in the vertical dimension.
Search methods: Unrestricted literature search of five databases up to May 2023.
Selection criteria: Randomized/non-randomized clinical studies comparing cHG to untreated controls, high-pull headgear (hp-HG), cHG ad-
juncts, or other Class II treatment alternatives (functional appliances or distalisers).
Data collection and analysis: After duplicate study selection, data extraction, and risk-of-bias assessment according to Cochrane, random-
effects meta-analyses of mean differences (MD)/standardized mean diffences (SMD) and their 95% confidence intervals (CIs) were performed,
followed by meta-regressions, sensitivity analyses, and assessment of certainty on existed evidence.
Results: Two randomized/16 non-randomized studies (12 retrospective/4 prospective) involving 1094 patients (mean age 10.9 years and 46%
male) were included. Compared to natural growth, cHG treatment was not associated on average with increases in mandibular (eight studies;
SMD 0.22; 95% CI −0.06, 0.49; P = 0.11) or maxillary plane angle (seven studies; SMD 0.81; 95% CI −0.34, 1.95; P=0.14). Observed changes
translate to MDs of 0.48° (95% CI −0.13, 1.07°) and 1.22° (95% CI −0.51, 2.94°) in the SN-ML and SN-NL angles, respectively. No significant
differences were seen in y-axis, facial axis angle, or posterior face height (P > 0.05). Similarly, no significant differences were found between
cHG treatment and (i) addition of a lower utility arch, (ii) hp-HG treatment, and (iii) removable functional appliance treatment (P > 0.05 for all).
Meta-regressions of patient age, sex, or duration and sensitivity analyses showed relative robustness, while our confidence in these estimates
was low to very low due to the risk of bias, inconsistency, and imprecision.
Conclusions: cHG on average is not consistently associated with posterior rotation of the jaws or a consistent increase in vertical facial dimen-
sions among Class II patients.
Registration: PROSPERO registration (CRD42022374603).
Keywords: orthodontics; Class II malocclusion; cervical pull headgear; clinical trials; systematic review; meta-analysis

Introduction low- and high-pull) HG. Several authors have reported dif-
Rationale ferential treatment effects according to the direction of the
applied force from HG: (i) cHG tends to extrude the maxil-
Since its introduction more than a century ago, the use of
lary dentition, rotate more the mandible backward more than
extraoral traction with headgear (HG) has gained a prominent
hp-HG [1, 3, 4], and might lead to open-bite [5]; (ii) hp-HG
place in orthodontic therapy for a wide spectrum of dental
leads to greater forward movement of the chin than cHG but
goals (including exerting influence on the sagittal or vertical
might not be as effective as the latter for severe protrusion
position of the upper molars and expanding the dental arch),
cases [3, 6, 7].
orthopaedic goals (retardation of maxillary growth) [1, 2] or
Furthermore, several authors have cautioned in the past
as a means to reinforce orthodontic anchorage.
against the use of cHG especially in dolichofacial patients,
Extraoral traction with HG is usually categorized according
as it might lead to molar extrusion, which in turn induces
to the direction of the applied force into cervical headgear
clockwise (backward) mandibular rotation and an increase
(cHG), high-pull headgear (hp-HG), or combination (both

© The Author(s) 2023. Published by Oxford University Press on behalf of the European Orthodontic Society
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/),
which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited.
2 European Journal of Orthodontics, 2024

in mandibular plane angle, thereby worsening a potentially review’s primary outcome was the inclination of the man-
already unattractive profile [1, 4, 8–10]. dibular plane assessed with the sela-nasion mandibular plane
On the other hand, such backward rotational effects have (SN-ML) or the Frankfort horizontal mandibular plane (FH-
been reported to be reversible, so that anterior growth of ML) angle. Secondary outcomes included (i) the inclination
both jaws is eventually seen and other factors such as occlusal of the maxillary plane assessed with the sela-nasion max-
forces or occlusal contacts might also influence the final out- illary plane (SN-NL) or the Frankfort horizontal maxillary
come of cHG treatment [9]. Moreover, other authors have plane (FH-NL) angle, (ii) the y-axis to the anterior cranial
reported that cHG did not cause more molar eruption than base (N-S-Gn), (iii) the facial axis angle (BaN-PtGn), (iv) the
would be expected from normal eruption [11] and did not lower posterior face height (Ar-Go), and the (v) total pos-
produce excessive backward rotation of the mandible [12, 13] terior face height (S-Go).
even for dolichofacial patients [11], while the vertical skeletal
relationships in the growing face could not be predictably al- Information sources and search strategy
tered by cHG treatment [14]. Finally, even if adverse effects An unrestricted literature search of five electronic databases
on the vertical dimension can be expected from cHG treat- (Medline via PubMed, Scopus, Web of Science, Cochrane
ment, some adjuncts like incorporation of a lower utility arch CENTRAL, and LILACS) was conducted from inception

Downloaded from https://academic.oup.com/ejo/article/46/1/cjad053/7326801 by guest on 06 May 2024

from Rickett’s bioprogressive therapy have been suggested to up to 1 May 2023, using an appropriate search strategy
minimize these [15]. (Supplementary Table 1). No restrictions regarding publica-
A previous systematic review on the subject [10] assessed tion date, language, type, or status were used, while the ref-
treatment effects solely from groups of patients treated with erence lists of eligible articles or previous systematic reviews
cHG and was limited to descriptive analysis without any were manually searched for any additional relevant articles.
proper quantitative synthesis (meta-analysis) of the relative
effects compared to normal growth, hp-HG, or other Class Selection process
II treatment alternatives, thereby failing to draw definite Initially, the titles and/or abstracts of all studies identified by
conclusions. the literature search were assessed against the eligibility cri-
teria, followed by retrieval and assessment of their full texts.
Objectives Study selection was performed independently by two authors
The aim of this systematic review was to assess clinical evi- (UM and FR) and any discrepancies were resolved by discus-
dence on the vertical effects of cHG in growing skeletal Class sion with a third author (AMS).
II patients. The review aimed to answer the following focussed
question: Does treatment with cHG for growing patients with Data collection process and items
Class II malocclusion have an effect on vertical cephalometric Data collection was performed using pre-defined and piloted
measurements of the craniofacial complex other than could extraction forms covering (i) study characteristics (design,
be expected from natural growth? Secondarily, the present re- clinical setting, and country); (ii) patient characteristics (age
view aimed to compare the effects of cHG to the use of any and sex); (iii) appliance characteristics; (iv) measured out-
other treatment adjuncts (like utility arches), hp-HG, or other comes; and (v) follow-up duration. To ensure accuracy and
treatment alternatives for Class II treatment (like functional consistency, all data was extracted independently by two au-
appliances or molar distalization appliances). thors (UH and FR), while any discrepancies were again re-
solved by discussion with a third author (AMS).

Materials and methods Study risk of bias

Registration and protocol The risk of bias of randomized trials was assessed with the
Cochrane risk of bias in randomized trials (RoB 2) tool [19].
This review was conducted according to the Cochrane hand-
The risk of bias within included non-randomized studies was
book [16] and reported according to the Preferred Reporting
assessed with the ROBINS-I (risk of bias in non-randomized
Items for Systematic Reviews and Meta-Analyses (PRISMA)
studies of interventions) tool [20]. The risk of bias was as-
2020 statement [17]. Its protocol was developed a priori
sessed independently by two authors (UM and FR) and any
following the corresponding PRISMA extension [18], pre-
discrepancies were resolved by consulting another author
registered (CRD42022374603), and all post hoc changes to
(AC).
the protocol were transparently reported (Supplement).
Effect measures and synthesis measures
Eligibility criteria Studies were considered eligible for pooling if similar par-
The eligibility criteria were developed based on the PICOS ticipants, interventions, and comparisons existed, and suf-
(participants, intervention, comparison, outcomes, and ficient data was reported. In case of missing/partial data
study design) principle; P: growing patients of any sex with provided, we tried to calculate the missing data ourselves
Class II malocclusion without any craniofacial syndrome or (Supplement). The mean difference (MD) with its 95% con-
anomalies; I: cHG alone or in combination with fixed ap- fidence interval (CI) was generally chosen as effect measure,
pliances; C: no treatment (observation of natural growth), while the standardized mean difference (SMD) was used to
cHG with any adjuncts, hp-HG, or any other intraoral ap- combine cephalometric variables measuring similar out-
pliance; O: vertical cephalometric measurements; and S: clin- comes (like SN-ML and FH-ML). As the effects of cHG
ical comparative studies, including randomized trials and were expected to vary according to the patient’s chrono-
prospective/retrospective cohort (before-and-after) studies. logical/skeletal age, sex, baseline skeletal configuration,
Excluded were case series (defined as studies with <10 pa- angulation of the HG’s inner/outer arms, and patient com-
tients), case reports, animal, and non-clinical studies. The pliance, a random-effects model was a priori deemed more
U. Hussain et al. 3

appropriate to calculate the average distribution of cHG Identification of studies via Identification of studies via other
effects across the various scenarios, based on clinical and databases and registers methods

statistical reasoning [21]. A restricted maximum likelihood

variance estimator was chosen, and the CIs were adjusted 1033 records identified electronically 4 records identified manually

Identification
with the Hartung–Knapp–Sidik–Jonkman method [22, 23].
Between-study heterogeneity was assessed through inspec-
296 duplicates were removed
tion of forest plots, the tau2 (absolute heterogeneity) or
the I2 statistic (relative heterogeneity; inconsistency) and
uncertainty intervals were calculated around them [24].
741 records were screened
Heterogeneity was assessed in absolute/relative terms,
based on its localization on the forest plot, its effect on the
summary estimate, and uncertainty around them. To appro- 634 were excluded by title/abstract
priately interpret the results of the random-effects model,
95% predictions were calculated to incorporate existing
heterogeneity and provide a range of possible effects for a 107 full texts were checked for eligibility against the criteria

Downloaded from https://academic.oup.com/ejo/article/46/1/cjad053/7326801 by guest on 06 May 2024

future clinical scenario [25]. Contour-enhanced forest plots

Screening
[26] were constructed to visualize the magnitude of ob-
1 Outcome not clear
served effects (Supplement) and assess heterogeneity, clin- 15 Case reports
ical relevance, and imprecision. 1 Inaccessible
1 Other language
Post-hoc random-effects meta-regressions were performed 3 Ineligible controls
18 ineligible outcomes
for meta-analyses with at least five studies to assess the im- 24 Duplicate publications
pact of patient age, % of male patients, and follow-up dur- 1 Unclear outcome
1 Unclear intervention
ation on the treatment results. Sensitivity analyses were 4 Co-interventions
performed again for meta-analyses with at least five studies 1 Irrelevant
1 Book chapter
to assess the impact of (i) study design (randomized or non- 15 Ineligible interventions
2 Review
randomized studies), (ii) timing of data acquisition (pro- 1 No Class II
spective or retrospective studies), (iii) sample size (up to or
Included

more than 50 patients per study; arbitrarily chosen), and (iv)

18 studies included in systematic review
our certainty on the meta-analytical estimates (quality of clin-
ical recommendations).
Statistical analyses were run in R statistical software (ver- Figure 1. PRISMA flow diagram for the identification and selection of
studies eligible for this review.
sion 4.0.4; R Foundation for Statistical Computing, Vienna,
Austria) by one author (SNP) with an openly provided dataset
[27]. All P-values are two‐sided with alpha = 0.05, except for Study characteristics
tests of between‐studies or between‐subgroups heterogeneity The characteristics of the 18 included studies are shown in
where alpha was set at 0.10. Table 1. The majority of studies (67%; 12/18) were retro-
spective non-randomized cohort (before-and-after) studies,
Reporting bias assessment and certainty 22% (4/18) were prospective non-randomized cohort
assessment (before-and-after) studies, and 11% (2/18) were randomized
Reporting biases (including small-study effects and the pos- clinical trials of parallel design. Included studies were con-
sibility of publication bias) were planned to be assessed for ducted in university clinics or private practices in 11 different
meta-analysis of at least 10 studies, but no such meta-analyses countries (Brazil, Denmark, Finland, Greece, Italy, Republic
could ultimately be done. of Korea, Spain, Sweden, Switzerland, Turkey, and USA).
Our certainty around the meta-analysis results was assessed These 18 studies included a total of 1094 patients (median
with the grades of recommendations, assessment, develop- 57 patients/study and 26 patients/study group), who were on
ment, and evaluation (GRADE) approach [28] and summar- average 45.7% male (423/925; from the 15 studies reporting
ized with revised summary of findings tables [29]. on sex) and on average 10.9 years old (from the 17 studies
reporting on age). Among the 18 included studies, more than
half of them (56%; 10/18) compared cHG to an untreated
Results control group, three of them (17%) to a cHG group with
a lower utility arch, five of them (26%) to a hp-HG group,
Study selection
four of them (21%) to an intraoral distaliser group, and
The electronic database search yielded a total of 1033 re- two of them (11%) to a functional appliance group. Most
cords, while another four were identified manually (Fig. studies included patients of any vertical skeletal configur-
1). After removal of 296 duplicates, 741 records remained ation, while four studies (22%) included only hyperdivergent
for further evaluation and were checked against the eligi- patients and another two studies (11%) included normo- or
bility criteria (Supplementary Table 2). Four publications hyperdivergent patients. The median treatment duration (and
from the same research team from Italy were identified [12, study follow-up) was 24.3 months.
30–32] and after communication with the authors were
grouped as a single clinical study with multiple compari-
sons. In the end, 21 publications pertaining to 18 unique Risk of bias in studies
clinical studies were included in the quantitative and quali- The risk of bias of the two included randomized trials is given
tative synthesis. in Supplementary Table 3 and Supplementary Fig. 1. Both
4 European Journal of Orthodontics, 2024

Table 1. Characteristics of included studies.

Nr Study Design; Groups Patients (M/F); age† Selected patients by Duration

setting* vertical classification (months)

1 Aliò-Sanz (2012) pNRS; ESP E: cHG E: 41 (20/21); NR Any 42.0

C: No Tx C: 38 (22/16); NR
2 Antonarakis rNRS; E: cHG (+FABOTH) E: 30 (15/15); 10.8 Any 46.8
(2014) CHE C: hp-HG (+FABOTH) C: 30 (15/15); 10.8
3 Bondermark RCT; SWE E: cHG E: 20 (8/12); 11.5 Any 6.4
(2005) C: Intraoral distaliser C: 20 (10/10); 11.4
4 Burke (1992) rNRS; USA E: cHG (+/−FABOTH) E: 21 (NR); 10.2 Hyperdivergent (SN-ML 43.3
C: hp-HG (+/−FABOTH) C: 32 (NR); 10.5 > 34º)
5 Cook (1994) rNRS; USA E1: cHG E1: 30 (14/16); 8.6 Any 19.0
E2: cHG + LUA E2: 30 (21/9); 8.7
C: No Tx C: 30 (15/15); 9.1

Downloaded from https://academic.oup.com/ejo/article/46/1/cjad053/7326801 by guest on 06 May 2024

6 Derringer (1990) rNRS; E: cHG E: 40 (20/20); 11.7 Any 51.1
DNK C1: Activator C1: 30 (15/15); 11.8
C2: No Tx C2: 22 (11/11); 11.7
7 Freitas (2008) rNRS; BRA E: cHG (+FAMND) E: 25 (5/20); 10.4 Any 30.0
C: No Tx C: 16 (4/12); 9.9
8 Gkantidis (2011) rNRS; E: cHG (+FABOTH) E: 28 (14/14); 11.0 Hyperdivergent (SN-ML 28.8
GRC C: hp-HG (+FABOTH + 4 PM-Ex) C: 29 (13/16); 11.8 > 32º)
9 Haralabakis rNRS; E: cHG E: 30 (9/21); 10.9 Any 27.5
(2003) GRC C: Activator C: 22 (11/11); 10.2
10 Kim (2000) rNRS; USA E: cHG (+FABOTH) E: 30 (7/23); 11.1 Any 49.1
C: No Tx C: 26 (10/16); 11.1
11 Lione (2014) rNRS; ITA E: cHG (+FABOTH) E: 40 (15/25); 11.5 Any 18.0
C1: Intraoral distaliser (+FABOTH) C1: 40 (19/21); 11.6
C2: No Tx C2: 25 (12/13); 11.4
12 Mantysaari RCT; FIN E: cHG E/C: 68 (40/28); 7.6 Any 16.0
(2004) C: No Tx
13 Mossaz (2005) pNRS; E: cHG (+FABOTH) E: 30 (NR); 11.7 Normo-/hyperdivergent 24.5
CHE C: Intraoral distaliser (+FABOTH) C: 30 (NR); 11.6 (NL-ML 20-35)
14 Park (2017) rNRS; E: cHG E: 22 (6/16); 23.0 Any 24.1
KOR C: Intraoral distaliser (skeletal) C: 22 (6/16); 24.7
15 Rosa (2020) pNRS; E: cHG E: 23 (10/13); 10.7 Any 15.0
BRA C: No Tx C: 22 (10/12); 10.7
16 Sambataro rNRS; ITA E1: cHG E1: 20 (10/10); 8.5 Hyperdivergent 21.6
(2017)collated E2: cHG + LUA E2: 19 (10/9); 8.6 (BaN-PtGn < 90º)
E3: hp-HG E3: 15 (NR); 9.4
C: No Tx C: 21 (11/10); 8.4
17 Ulger 2006 pNRS; E1: cHG E1: 12 (6/6): 8.9 Normo-/hyperdivergent 17.2
TUR E2: cHG + LUA E2: 12 (5/7); 9.2
C: No Tx C: 12 (4/8); 8.6
18 Zervas (2016) rNRS; USA E: cHG E: 22 (NR); 8.6 Hyperdivergent (BaN- 10.0
C: hp-HG C: 19 (NR); 9.4 PtGn < 90º; TFH > 57º)

*
Countries are given with their ISO ALPHA-3 codes.
†
In years.
cHG, cervical headgear; C, control group; E, experimental group; FABOTH, Fixed appliance on both jaws; FAMAX, Fixed appliance on the upper jaw; FAMND,
Fixed appliance on the lower jaw; hp-HG, high-pull headgear; LUA, lower utility arch; NR, Not reported; PM-Ex; premolar extraction; pNRS, prospective
non-randomized study; RCT, Randomised clinical trial; rNRS, retrospective nonrandomised study; TFH, total facial height; Tx, treatment.

randomized trials were judged to be in high risk of bias, due to Results of individual studies and data syntheses
issues with the randomization process, deviations from intended The complete results extracted from all included studies can
interventions, and measurement of the outcome of interest. be found in the review’s openly provided dataset [27]. Results
The risk of bias of the 16 included non-randomized studies is of meta-analyses with at least two studies can be seen in Table
given in Supplementary Table 4 and Supplementary Figs. 2 and 2, while outcomes/comparisons assessed only from single
3. From these, half of them (50%; 8/16) were judged to be in studies can be seen in Supplementary Table 5—the latter not
moderate risk of bias and the other half (50%; 8/16) in high risk finding any clinically relevant differences between cHG and
of bias. The most problematic domains were bias due to con- untreated controls, hp-HG, addition of a lower utility arch to
founding, bias due to the selection of the study’s participants, the cHG, intraoral distaliser, functional appliance, or intru-
and bias due to deviations from the intended interventions. sive mechanics.
U. Hussain et al. 5

As far as comparisons of cHG to natural growth (untreated significant difference in mandibular plane angle was seen be-
controls) are concerned, random-effects meta-analyses in- tween treatment with cHG and functional appliance (two
dicated that cHG was not associated with significantly in- studies; P = 0.85; Supplementary Fig. 15).
creased mandibular plane angle (measured with SN-ML or Meta-regression analyses found no significant effect of
FH-ML) (Table 3). Pooling the results of eight studies (Fig. baseline patient age, % of male patients within the study
2), an SMD of 0.22 was found (95% CI -0.06 to 0.49; P = sample, or follow-up duration on the effects of cHG com-
0.11), which indicated a moderate effect on average and can pared to untreated controls—in terms of mandibular or max-
be back-translated in SN-ML to an increase by 0.48º (95% CI illary plane angle (P > 0.10 in most instances; Supplementary
-0.13 to 1.07º.). This is not statistically significant and is less Table 6). The only exception was patient sex, where male pa-
than half the average baseline standard deviation for SN-ML tients were associated with smaller opening of the mandibular
of the control group, which means it is surely of little clin- plane angle, which was however of very small magnitude.
ical relevance. Other than that, meta-analyses indicated no Finally, sensitivity analyses found no significant differences
differences regarding maxillary plane inclination (through according to study design (randomized vs non-randomized
SN-NL or FH-NL; seven studies; P = 0.14; Supplementary studies), data acquisition timing (prospective vs pro-
Fig. 4), y-axis (N-S-Gn; two studies; P = 0.34; Supplementary spective studies), or study sample size (up to vs more than

Downloaded from https://academic.oup.com/ejo/article/46/1/cjad053/7326801 by guest on 06 May 2024

Fig. 5), facial axis angle (BaN-PtGn; three studies; P = 0.43; 50 patients/study), which indicated robustness of the results
Supplementary Fig. 6), lower posterior face height (Ar-Go; (Supplementary Table 7). The only exception was sensitivity
three studies; P = 0.15; Supplementary Fig. 7), or total pos- analysis by study design, where considerable greater increase
terior face height (S-Go; three studies; P = 0.32; Fig. 3). in mandibular plane angle was seen in weaker retrospective
Compared to hp-HG (Table 2), treatment with cHG was studies, but not in prospective studies.
not associated with increased mandibular plane angle (SN-
ML or FH-ML; four studies; P = 0.96; Supplementary Fig. 8), Certainty of evidence
maxillary plane angle (SN-NL or FH-NL; three studies; P = Our certainty on the results of the meta-analysis was assessed
0.32; Supplementary Fig. 9), or facial axis angle (BaN-PtGn; with the GRADE approach and was judged as either low or
three studies; P = 0.49; Supplementary Fig. 10). very low in all instances. The greatest issue was the high risk
Addition of a lower utility arch in the cHG protocol had of bias due to the inclusion of non-randomized studies with
similarly no effect on mandibular plane (SN-ML or FH-ML; methodological weaknesses. Furthermore, inconsistency was
three studies; P = 0.58; Supplementary Fig. 11) or maxillary found on the effect of cHG on mandibular plane angle, as
plane angle (SN-NL; three studies; P = 0.41; Supplementary the effect magnitude was unclear and reported effects ranged
Fig. 12). Likewise, no significant differences were seen between from small to moderate or large (Fig. 2). Furthermore, signs
treatment with cHG and an intraoral distaliser in terms of of inconsistency were seen in the effect of cHG on maxil-
mandibular (two studies; P = 0.36; Supplementary Fig. 13) or lary plane angle (Supplementary Fig. 4) that was overall not
maxillary plane angle (two studies; P = 0.94; Supplementary significant (P = 0.14) and was very heterogeneous (τ2 1.11;
Fig. 14). From these two studies, only one was on growing I2 80%). This was due to the outlier of the Ülger et al. [33]
patients and the other was on adult patients (Supplement), study that showed a much larger effect than all other studies.
but both found no significance difference on maxillary plane Omitting this study led to a much more precise meta-analysis
inclination (Supplementary Fig. 14)—the only outcome used that indicated increased maxillary plane inclination with cHG
from the single included study on adult patients. Finally, no (five studies; SMD 0.39; 95% CI 0.13 to 0.66; P = 0.01) that

Table 2. Results of meta-analyses (≥2 studies) comparing cervical headgear with other treatment alternatives.

Comparison Outcome n Effect (95% CI) P τ2 (95% UI) I2 (95% UI) 95% prediction

cHG vs control SN-ML/FH-ML 8 SMD 0.22 (−0.06, 0.49) 0.11 0.03 (0, 0.34) 23% (0%, 65%) −0.31, 0.74
SN-NL/FH-NL 7 SMD 0.81 (−0.34, 1.95) 0.14 1.11 (0.37, 8.90) 80% (60%, 90%) −2.11, 3.72
NSGn 2 MD 1.06 (−6.88, 9.00) 0.34 0.63 (NC) 80% (NC) NC
BaN-PtGn 3 MD 0.15 (−0.50, 0.79) 0.43 0 (0, 2.21) 0% (0%, 90%) −2.52, 2.82
ArGo 3 MD 2.31 (−2.01, 6.64) 0.15 2.43 (0.25, >100) 82% (43%, 94%) −21.18, 25.81
SGo 3 MD 1.30 (−2.99, 5.58) 0.32 2.47 (0.20, >100) 81% (39%, 94%) −22.65, 25.25
cHG vs hp-HG SN-ML/FH-ML 4 SMD −0.01 (−0.80, 0.77) 0.96 0.15 (0, 3.29) 63% (0%, 88%) −2.00, 1.97
SN-NL/FH-NL 3 SMD 0.51 (−1.15, 2.16) 0.32 0.35 (0.03, 17.17) 79% (34%, 94%) −8.51, 9.53
BaN-PtGn 3 MD −0.76 (−4.70, 3.18) 0.49 2.31 (0.48, >100) 92% (80%, 97%) −23.26, 21,74
cHG vs cHG + lower utility arch SN-ML/FH-ML 3 SMD 0.08 (−0.47, 0.64) 0.58 0 (0, 1.64) 0% (0%, 90%) −2.21, 2.38
SN-NL/FH-NL 3 SMD 0.18 (−0.57, 0.94) 0.41 0 (0, 3.73) 0% (0%, 90%) −2.14, 2.50
BaN-PtGn 2 MD −0.33 (−1.59, 0.93) 0.19 0 (NC) 0% (NC) NC
cHG vs intraoral distaliser SN-ML 2 MD −0.43 (−3.83, 2.97) 0.36 0.05 (NC) 22% (NC) NC
SN-NL/FH-NL 2 SMD 0.02 (−3.26, 3.31) 0.94 0.07 (NC) 48% (NC) NC
cHG vs functional appliance SN-ML 2 MD −0.08 (−4.44, 4.28) 0.85 0 (NC) 0% (NC) NC

cHG, cervical headgear; CI, confidence interval; hp-HG, high-pull headgear; MD, mean difference; NC, not calculable; SMD, standardized mean difference;
UI, uncertainty interval.
6 European Journal of Orthodontics, 2024

Table 3. Summary of findings table according to the GRADE approach.

Anticipated absolute effects (95%

CI)

Outcome studies Control Difference in cHG Quality of What happens Comment

(patients) groupa group the evidence with experimental
(GRADE)b treatment

cHG vs control (no Tx)

 Mandibular plane in- −0.33º 0.48º greater (0.13º ⨁◯◯◯ Little to no difference Based on an SMD for SN-ML/FH-ML
clination (SN-ML) lower to 1.07º Very lowc,d due to in mandibular plane of 0.22 (95% CI -0.06 to 0.49);
8 studies (394 patients) greater) bias, inconsist- inclination back-translated to SN-ML using an
ency average control SD of 2.19º.
 Maxillary plane in- +0.16º 1.22º greater (0.51º ⨁⨁◯◯ Little to no difference Based on an SMD for SN-NL/FH-NL
clination (SN-NL) lower to 2.94º Lowc,e due to bias in maxillary plane of 0.81 (95% CI -0.34 to 1.95);

Downloaded from https://academic.oup.com/ejo/article/46/1/cjad053/7326801 by guest on 06 May 2024

7 studies (391 studies) greater) inclination back-translated to SN-ML using an
average control SD of 1.51º.
 Y-axis (N-S-Gn) −0.02º 1.06º greater (6.88º ⨁◯◯◯ Little to no difference -
2 studies (118 patients) smaller to 9.00º Very lowc,f due to in y-axis
greater) bias, impreci-
sion
 Facial axis angle −0.23º 0.15º greater (0.50º ⨁◯◯◯ Little to no difference -
(BaN-PtGn) smaller to 0.79º Very lowc,f due to in facial axis
3 studies (161 patients) greater) bias, impreci-
sion
 Posterior face height +3.94 mm 1.30 mm greater ⨁⨁◯◯ Little to no difference -
(S-Go) (2.99 mm smaller Lowc due to bias in posterior face
3 studies (121 patients) to 5.58 mm height
greater)
cHG vs hp-HG
 Mandibular plane in- +0.83º 0.02º smaller (1.90º ⨁◯◯◯ Little to no difference Based on an SMD for SN-ML/FH-ML
clination (SN-ML) smaller to 1.83º Very lowc,f due to in mandibular plane of −0.01 (95% CI -0.80 to 0.77);
4 studies (189 patients) greater) bias, impreci- inclination back-translated to SN-ML using an
sion average control SD of 2.38º.
 Maxillary plane in- 0º 1.12º greater (2.53º ⨁◯◯◯ Little to no difference Based on an SMD for SN-NL/FH-NL
clination (SN-NL) smaller to 4.75º Very lowc,f due to in maxillary plane of 0.51 (95% CI -1.15 to 2.16);
3 studies (136 studies) greater) bias, impreci- inclination back-translated to SN-ML using an
sion average control SD of 2.20º.
 Facial axis angle −0.06º 0.76º smaller (4.70º ⨁◯◯◯ Little to no difference –
(BaN-PtGn) smaller to 3.18º Very lowc,f due to in facial axis angle
3 studies (129 patients) greater) bias, impreci-
sion

Population: skeletal class II malocclusion; intervention: cervical headgear (+/− braces); comparison: no treatment (control) or high-pull headgear; setting:
university clinics or private practices (Brazil, Denmark, Finland, Greece, Italy, Republic of Korea, Spain, Sweden, Switzerland, Turkey, and USA).
a
Response in the control group is based on the response of included studies (or random-effects meta-analysis of the control response).
b
Starts from ‘high’.
c
Downgraded by two levels, due to serious potential issues with confounding, selection of participants, and deviation of intended intervention.
d
Signs of inconsistency, as potential effects include small reductions to very large increases.
e
Potential for inconsistency, as the CIs/prediction included a wide range of outcomes. However, this was mostly due to a very heterogeneous study (Ulger
2006) with a very large effect size. Omission of this study led to much more precise estimates (SMD 0.39; 95% CI 0.13 to 0.66; P = 0.01). Decided not to
downgrade.
f
Imprecision due to the limited number of small studies.
cHG, cervical headgear; CI, confidence interval; hp-HG, high-pull headgear; SD, standard deviation; SMD, standardized mean difference; Tx, treatment.

might be more appropriate than the original analysis. This treatment alternatives and is to the best of our knowledge the
can be back-translated to an increase in SN-NL by 0.59º first study of its kind.
(95% CI 0.20 to 1.00º), which even though statistically sig- The results of the meta-analyses indicated that treatment
nificant, is of little clinical relevance. with cHG was associated with a minimal non-significant pos-
terior rotation of the mandible and the maxilla compared to
natural growth (0.48º and 1.22º, respectively), which is how-
Discussion ever of little clinical relevance. This comes in contrast with a
previous narrative analysis of the literature [10] that reported
Results in context bite opening and increased vertical cephalometric measure-
The present review systematically appraised evidence from 18 ments after cHG treatment. This also contradicts previous
clinical studies and a total 1094 patients being treated with opinions that cHG is de facto contraindicated for high-angle
cHG and compared to untreated controls or other Class II facial types due to its clockwise (backward) mandibular
U. Hussain et al. 7

Study
small moderate

SN-ML / FH-ML
large very large

SMD 95%-CI Weight

in patients with strong musculature. However, the utility arch
was eventually effective in tipping the lower molar distally
Derringer 1990 -0.23 [-0.75; 0.29] 14.4%
Cook 1994 0.02 [-0.49; 0.53] 15.0% and maintaining its mesiodistal position but did not have a
Rosa 2020 0.05 [-0.53; 0.63] 12.3% significant effect on its vertical position [11, 33].
Ulger 2006 0.18 [-0.62; 0.98] 7.4%
Sambataro 2017col 0.23 [-0.38; 0.85] 11.4% Class II treatment with cHG was found to have similar ef-
Lione 2015
Freitas 2008
0.24
0.50
[-0.26; 0.74]
[-0.13; 1.14]
15.2%
10.8%
fects in the vertical direction with removable functional ap-
Kim 2001 0.81 [ 0.26; 1.36] 13.5% pliances both in terms of mandibular plane angle (Table 2;
RE model (HK) 0.22 [-0.06; 0.49] 100.0%
Supplementary Fig. 15) and in terms of maxillary plane angle,
Prediction interval [-0.31; 0.74] y-axis, and lower posterior face height (Supplementary Table
Heterogeneity: I 2 = 23% -1.6 -0.8 -0.5 -0.2 0 0.2 0.5 0.8 1.6 5). This is in agreement with the observation of Baumrind
et al. [39] who reported that treatment with cHG was as-
Figure 2. Contour-enhanced forest plot for the effect of cervical headgear
versus control (no treatment) on mandibular plane angle (SN-ML/FH-ML). sociated with an increase in lower face height compared to
CI, confidence interval; SMD, standardized mean difference. natural growth or hp-HG and which was similar to treatment
with Activator, but at the same time additional growth at the
ramus kept the mandibular plane angle stable. Removable

Downloaded from https://academic.oup.com/ejo/article/46/1/cjad053/7326801 by guest on 06 May 2024

small moderate large very large

Study S-Go (mm) MD 95%-CI Weight functional appliances are known to induce a small increase
in mandibular plane angle, which amount to about 0.66⁰
Kim 2001 -0.30 [ -2.54; 1.94] 28.7%
Freitas 2008 0.57 [ -1.57; 2.71] 29.7%
increase in SN-ML per treatment year compared to natural
Ulger 2006 2.92 [ 2.18; 3.66] 41.6% growth, but no consistent rotational effect on the maxilla
[40]. Generally, however, it seems that treatment-related ef-
RE model (HK) 1.30 [ -2.99; 5.58] 100.0% fects from usual Class II correction methods on mandibular
Prediction [-22.65; 25.25]
plane orientation are too small in themselves to be of major
I2=80%: -25 -10 -5 -2.5 0 2.5 5 10 25
clinical relevance [41, 34] and that cHG treatment simply cir-
Figure 3. Contour-enhanced forest plot for the effect of cervical headgear cumvents the usual reduction in mandibular pane angle of
versus control (no treatment) on posterior face height (S-Go). CI, normal growth [42].
confidence interval; MD, mean difference. Finally, no significant difference in vertical effects was found
between cHG and dentally anchored intraoral distalisers. This
is logical, as mostly similar extrusive effects for the maxillary
rotation and increase in mandibular plane angle that can molar were seen between the two groups [43, 44].
worsen a potentially already unattractive profile [1, 4, 8–10]
and hp-HG being a more appropriate choice for such cases [1, Strengths and limitations
35, 36] Possible explanations for this include among others This review has several strengths including a priori registra-
that the HG-induced molar extrusion is offset by a significant tion [45], an extensive unrestricted literature search, robust
increase in ramus height due to increased condylar growth analytical methods [22], sensitivity analyses to check the
[11, 12, 33] and therefore no significant increase in man- influence of methodological characteristics on the studies’
dibular plane angle is seen. results, its transparent open data availability [46], and assess-
Similarly, the present review failed to find that treatment ment of our confidence in the meta-analysis results through
with cHG resulted in more pronounced backward growth ro- the GRADE approach.
tation, since no difference in the y-axis or the facial axis angle However, certain limitations exist for this review. First and
was found with either untreated controls or hp-HG (Table foremost, most of the included studies were non-randomized,
2). This is in agreement with the notion by Melsen [9] who many were retrospective, and some also included historical
found that both cHG and hp-HG had a similar effect on the control groups—study design characteristics that have all
growth direction of the maxilla or mandible and a mostly an- been linked to increased risk of bias [47–49]. Furthermore,
teriorly directed mandibular growth is seen after both treat- information like baseline skeletal configuration as selection
ment alternatives. criterion, the vertical angulation/length of the external HG
It has been proposed that cHG is associated with increased bows, calculated line of applied force according to the centre
height at the maxillary molar that mimics natural growth, of resistance, magnitude of applied forces, and compliance
produces occlusal interferences, and subsequently leads to with prescribed wear might influence the observed treatment
forward movement of the mandible to maintain the occlusal effects [3, 9, 50–55], but were not adequately reported in in-
contacts [7, 37]. Indeed, this slightly greater extrusion of the cluded studies and could therefore not be formally assessed
maxillary molar seemed to be compensated by a positional statistically in this review that provides the average distribu-
stability of the lower molar that extruded significantly less tion of cHG effects.
after treatment with cHG than with hp-HG [7] and agrees
with previous reports [38].
Combination of cHG with a lower utility arch was not found Conclusions
from the present review to be associated with significantly Based on available evidence from mostly non-randomized
different vertical effects than treatment cHG alone. Ricketts clinical studies assessing the effect of Class II treatment with
had propagated that the addition of a utility arch to cHG cHG, mostly minor effects on vertical parameters were seen.
avoids incisor interference and through this reverse response Compared to natural growth cHG treatment was not con-
prevents opening rotation of the mandible [12]. Theoretically, sistently associated with increases in the maxillary and man-
use of a lower utility arch would lead to a stabilizing reverse dibular plane angles, while no effects on posterior face height
response for the mandibular plane angle and the facial axis or growth direction were seen. No considerable differences on
angle [15] or even to a counterclockwise (forward) rotation the vertical effects were seen between cHG and addition of a
8 European Journal of Orthodontics, 2024

lower utility arch, hp-HG, functional appliances, or intraoral 7. Zervas ED, Galang-Boquiren MTS, Obrez A et al. Change in the
distalisers. However, our certainty about these findings is vertical dimension of Class II Division 1 patients after use of cer-
limited due to serious methodological limitations of the cur- vical or high-pull headgear. Am J Orthod Dentofacial Orthop
rent evidence base and future studies with more robust design 2016;150:771–81. https://doi.org/10.1016/j.ajodo.2016.04.024
8. Creekmore TD. Inhibition or stimulation of the vertical
might shed more light on this matter.
growth of the facial complex, its significance to treatment.
Angle Orthod 1967;37:285–97. https://doi.org/10.1043/0003-
3219(1967)037<0285:IOSOTV>2.0.CO;2
Acknowledgements
9. Melsen B. Effects of cervical anchorage during and after treat-
We thank Salvo Bocchieri (University of Insubria) for provid- ment: an implant study. Am J Orthod 1978;73:526–40. https://doi.
ing clarifications on studies from their research group. org/10.1016/0002-9416(78)90242-7
10. Henriques FP, Janson G, Henriques JFC et al. Effects of cervical
headgear appliance: a systematic review. Dental Press J Orthod
Author contributions 2015;20:76–81. https://doi.org/10.1590/2176-9451.20.4.076-081.
oar
Umar Hussain (Conceptualization [equal], Investigation
11. Cook AH, Sellke TA, BeGole EA. Control of the vertical dimension
[equal], Methodology [equal], Project administration [equal],

Downloaded from https://academic.oup.com/ejo/article/46/1/cjad053/7326801 by guest on 06 May 2024

in Class II correction using a cervical headgear and lower utility
Supervision [equal], Writing—original draft [equal]), Ahsan arch in growing patients. Part I. Am J Orthod Dentofacial Orthop
Shah (Conceptualization [equal], Investigation [equal], 1994;106:376–88. https://doi.org/10.1016/S0889-5406(94)70059-1
Methodology [equal], Writing—review & editing [equal]), 12. Sambataro S, Fastuca R, Oppermann NJ et al. Cephalometric
Fazli Rabi (Conceptualization [equal], Investigation [equal], changes in growing patients with increased vertical dimension
Writing—review & editing [equal]), Alessandra Campobasso treated with cervical headgear. J Orofac Orthop 2017;78:312–20.
(Conceptualization [equal], Investigation [equal], https://doi.org/10.1007/s00056-017-0087-z
Methodology [equal], Writing—review & editing [equal]), 13. Gkantidis N, Halazonetis DJ, Alexandropoulos E et al. Treat-
and Spyridon Papageorgiou (Data curation [equal], Formal ment strategies for patients with hyperdivergent Class II Divi-
sion 1 malocclusion: is vertical dimension affected? Am J Orthod
analysis [equal], Investigation [equal], Software [equal],
Dentofacial Orthop 2011;140:346–55. https://doi.org/10.1016/j.
Supervision [equal], Writing—review & editing [equal]). ajodo.2011.05.015
14. Haralabakis NB, Sifakakis IB. The effect of cervical headgear on
patients with high or low mandibular plane angles and the ‘myth’ of
Conflict of interest posterior mandibular rotation. Am J Orthod Dentofacial Orthop
The authors declare that there is no conflict of interest. 2004;126:310–7. https://doi.org/10.1016/j.ajodo.2003.08.028
15. Ricketts RM. Bioprogressive therapy as an answer to ortho-
dontic needs. part I. Am J Orthod 1976;70:241–68. https://doi.
Funding org/10.1016/0002-9416(76)90332-8
None declared. 16. Higgins JPT, Thomas J, Chandler J et al. (eds). (2020) Cochrane
Handbook for Systematic Reviews of Interventions Version 6.1
(Updated September 2020). Cochrane. www.training.cochrane.org/
Data availability handbook (9 September 2023, date last accessed).
17. Page MJ, Moher D, Bossuyt PM et al. PRISMA 2020 explanation
The data underlying this article are openly available at and elaboration: updated guidance and exemplars for reporting
Zenodo: http://doi.org/10.5281/zenodo.8173798 (27). systematic reviews. Br Med J 2021;372.
18. Shamseer L, Moher D, Clarke M et al. Preferred reporting items for
systematic review and meta-analysis protocols (PRISMA-P) 2015:
Supplementary material elaboration and explanation. Br Med J 2015;350:g7647. https://
Supplementary material is available at European Journal of doi.org/10.1136/bmj.g7647
19. Sterne JAC, Savović J, Page MJ et al. RoB 2: a revised tool for
Orthodontics online.
assessing risk of bias in randomised trials. Br Med J 2019;366:l4898.
https://doi.org/10.1136/bmj.l4898
REFERENCES 20. Sterne JA, Hernán MA, Reeves BC et al. ROBINS-I: a tool for
assessing risk of bias in non-randomised studies of interventions.
1. Merrifield LL, Cross JJ. Directional forces. Am J Orthod Br Med J 2016;355:i4919. https://doi.org/10.1136/bmj.i4919
1970;57:435–64. https://doi.org/10.1016/0002-9416(70)90164-8 21. Papageorgiou SN. Meta-analysis for orthodontists: part I–how to
2. Papageorgiou SN, Kutschera E, Memmert S et al. Effectiveness of choose effect measure and statistical model. J Orthod 2014;41:317–
early orthopaedic treatment with headgear: a systematic review 26. https://doi.org/10.1179/1465313314Y.0000000111
and meta-analysis. Eur J Orthod 2017;39:176–87. https://doi. 22. Langan D, Higgins JP, Jackson D et al. A comparison of het-
org/10.1093/ejo/cjw041 erogeneity variance estimators in simulated random‐effects
3. Barton JJ. High-pull headgear versus cervical traction: a cephalo- meta‐analyses. Res Synth Methods 2019;10:83–98. https://doi.
metric comparison. Am J Orthod 1972;62:517–29. https://doi. org/10.1002/jrsm.1316
org/10.1016/0002-9416(72)90027-9 23. van Aert RCM, Jackson D. A new justification of the Hartung-
4. Brown P. A cephalometric evaluation of high-pull molar headgear Knapp method for random-effects meta-analysis based on weighted
and face-bow neck strap therapy. Am J Orthod 1978;74:621–32. least squares regression. Res Synth Methods 2019;10:515–27.
https://doi.org/10.1016/0002-9416(78)90002-7 https://doi.org/10.1002/jrsm.1356
5. Watson WG. A computerized appraisal of the high-pull face-bow. 24. Higgins JP. Commentary: heterogeneity in meta-analysis should
Am J Orthod 1972;62:561–79. https://doi.org/10.1016/0002- be expected and appropriately quantified. Int J Epidemiol
9416(72)90001-2 2008;37:1158–60. https://doi.org/10.1093/ije/dyn204
6. Kuhn RJ. Control of anterior vertical dimension and proper selec- 25. IntHout J, Ioannidis JP, Rovers MM et al. Plea for routinely
tion of extraoral anchorage. Angle Orthod 1968;38:340–9. https:// presenting prediction intervals in meta-analysis. BMJ Open
doi.org/10.1043/0003-3219(1968)038<0340:COAVDA>2.0.CO;2 2016;6:e010247. https://doi.org/10.1136/bmjopen-2015-010247
U. Hussain et al. 9

26. Papageorgiou SN. Meta-analysis for orthodontists: part II—Is all 41. Baumrind S, Molthen R, West EE et al. Mandibular plane changes
that glitters gold? J Orthod 2014;41:327–36. https://doi.org/10.11 during maxillary retraction. Am J Orthod 1978;74:32–40. https://
79/1465313314Y.0000000110 doi.org/10.1016/0002-9416(78)90043-x
27. Hussain U, Shah AM, Rabi F et al. Vertical effects of cervical cer- 42. Hubbard GW, Nanda RS, Currier GF. A cephalometric eval-
vical headgear in growing patients with Class II malocclusion: a uation of nonextraction cervical headgear treatment in Class
systematic review and meta-analysis [Data set]. Zenodo 2023. II malocclusions. Angle Orthod 1994;64:359–70. https://doi.
https://doi.org/10.5281/zenodo.8173798 org/10.1043/0003-3219(1994)064<0359:ACEONC>2.0.CO;2
28. Guyatt GH, Oxman AD, Schünemann HJ et al. GRADE guidelines: 43. Bondemark L, Karlsson I. Extraoral vs intraoral appliance for distal
a new series of articles in the Journal of Clinical Epidemiology. movement of maxillary first molars: a randomized controlled trial.
J Clin Epidemiol 2011;64:380–2. https://doi.org/10.1016/j. Angle Orthod 2005;75:699–706. https://doi.org/10.1043/0003-
jclinepi.2010.09.011 3219(2005)75[699:EVIAFD]2.0.CO;2
29. Carrasco-Labra A, Brignardello-Petersen R, Santesso N et al. 44. Mossaz CF, Byloff FK, Kiliaridis S. Cervical headgear vs pendulum
Improving GRADE evidence tables part 1: a randomized trial appliance for the treatment of moderate skeletal Class II malocclu-
shows improved understanding of content in summary of findings sion. Am J Orthod Dentofacial Orthop 2007;132:616–23. https://
tables with a new format. J Clin Epidemiol 2016;74:7–18. https:// doi.org/10.1016/j.ajodo.2005.11.043
doi.org/10.1016/j.jclinepi.2015.12.007 45. Sideri S, Papageorgiou SN, Eliades T. Registration in the inter-

Downloaded from https://academic.oup.com/ejo/article/46/1/cjad053/7326801 by guest on 06 May 2024

30. Sambataro S, Lorusso P, Caprioglio A et al. Changes of occlusal national prospective register of systematic reviews (PROSPERO)
plane in growing patients with increased vertical dimension during of systematic review protocols was associated with increased re-
Class II correction by using cervical headgear. J Craniofac Surg view quality. J Clin Epidemiol 2018;100:103–10. https://doi.
2020;31:172–7. https://doi.org/10.1097/SCS.0000000000006197 org/10.1016/j.jclinepi.2018.01.003
31. Sambataro S, Bocchieri S, Fastuca R et al. Occlusal plane and skel- 46. Papageorgiou SN, Cobourne MT. Data sharing in orthodontic re-
etal changes after cervical headgear treatment with and without search. J Orthod 2018;45:1–3. https://doi.org/10.1080/14653125.
lower utility arch in Class II growing patients. J Craniofac Surg 2018.1440792
2021;32:1152–6. https://doi.org/10.1097/SCS.0000000000007305 47. Papageorgiou SN, Xavier GM, Cobourne MT. Basic study de-
32. Sambataro S, Rossi O, Bocchieri S et al. Comparison of cephalometric sign influences the results of orthodontic clinical investigations.
changes in Class II growing patients with increased vertical dimen- J Clin Epidemiol 2015;68:1512–22. https://doi.org/10.1016/j.
sion after high-pull and cervical headgear treatment. Eur J Paediatr jclinepi.2015.03.008
Dent 2023;24:36–41. https://doi.org/10.23804/ejpd.2023.24.01.06 48. Papageorgiou SN, Kloukos D, Petridis H et al. Publication of sta-
33. Ülger G, Arun T, Sayınsu K et al. The role of cervical headgear tistically significant research findings in prosthodontics & im-
and lower utility arch in the control of the vertical dimension. plant dentistry in the context of other dental specialties. J Dent
Am J Orthod Dentofacial Orthop 2006;130:492–501. https://doi. 2015;43:1195–202. https://doi.org/10.1016/j.jdent.2015.08.005
org/10.1016/j.ajodo.2005.01.027 49. Papageorgiou SN, Koretsi V, Jäger A. Bias from historical control
34. Haralabakis NB, Halazonetis DJ, Sifakakis IB. Activator versus groups used in orthodontic research: a meta-epidemiological study.
cervical headgear: superimpositional cephalometric comparison. Eur J Orthod 2017;39:98–105. https://doi.org/10.1093/ejo/cjw035
Am J Orthod Dentofacial Orthop 2003;123:296–305. https://doi. 50. Greenspan RA. Reference charts for controlled extraoral force
org/10.1067/mod.2003.20 application to maxillary molars. Am J Orthod 1970;58:486–91.
35. Poulton DR. The influence of extraoral traction. Am J Orthod https://doi.org/10.1016/0002-9416(70)90025-4
1967;53:8–18. https://doi.org/10.1016/0002-9416(67)90135-2 51. Huanca Ghislanzoni L, Ameur S, Antonarakis GS et al. Headgear
36. Gautam P, Valiathan A, Adhikari R. Craniofacial displacement in compliance as assessed by a temperature-sensitive recording device:
response to varying headgear forces evaluated biomechanically a prospective clinical study. Eur J Orthod 2019;41:641–5. https://
with finite element analysis. Am J Orthod Dentofacial Orthop doi.org/10.1093/ejo/cjz036
2009;135:507–15. https://doi.org/10.1016/j.ajodo.2007.02.059 52. Julku J, Pirilä-Parkkinen K, Tolvanen M et al. Comparison of
37. Björk A, Skieller V. Facial development and tooth eruption. an im- effects of cervical headgear treatment on skeletal facial changes
plant study at the age of puberty. Am J Orthod 1972;62:339–83. when the treatment time is altered: a randomized controlled trial.
https://doi.org/10.1016/s0002-9416(72)90277-1 Eur J Orthod 2019;41:631–40. https://doi.org/10.1093/ejo/cjz053
38. Burke M, Jacobson A. Vertical changes in high-angle Class II, di- 53. Talvitie TH, Helminen M, Karsila S et al. Impact of force magni-
vision 1 patients treated with cervical or occipital pull headgear. tude on effectiveness in cervical headgear therapy: a cephalometric
Am J Orthod Dentofacial Orthop 1992;102:501–8. https://doi. analysis. Eur J Orthod 2019;41:646–51. https://doi.org/10.1093/
org/10.1016/0889-5406(92)70066-J ejo/cjz022
39. Baumrind S, Korn EL, Molthen R et al. Changes in facial 54. Kallunki J, Bondemark L, Paulsson L. Early headgear acti-
dimensions associated with the use of forces to retract the max- vator treatment of Class II malocclusion with excessive overjet:
illa. Am J Orthod 1981;80:17–30. https://doi.org/10.1016/0002- a randomized controlled trial. Eur J Orthod 2021;43:639–47.
9416(81)90193-7 https://doi.org/10.1093/ejo/cjaa073
40. Koretsi V, Zymperdikas VF, Papageorgiou SN et al. Treatment 55. Talvitie T, Helminen M, Karsila S et al. The impact of force magni-
effects of removable functional appliances in patients with Class II tude on the first and second maxillary molars in cervical headgear
malocclusion: a systematic review and meta-analysis. Eur J Orthod therapy. Eur J Orthod 2021;43:648–57. https://doi.org/10.1093/
2015;37:418–34. https://doi.org/10.1093/ejo/cju071 ejo/cjab010