European Journal of Orthodontics 30 (2008) 120–127
Advance Access publication 23 January 2008
© The Author 2008. Published by Oxford University Press on behalf of the European Orthodontic Society.
All rights reserved. For permissions, please email: email@example.com.
The effi cacy of the Herbst appliance in normalizing the
sagittal dental arch relationship in patients with a Class II
division I malocclusion is well documented. The correction
is partly achieved by anterior mandibular tooth movement
and incisor proclination, as well as by posterior movement
of the maxillary teeth. The use of the Herbst appliance also
results in increased mandibular condylar growth ( Pancherz,
1979 , 1982 ; Paulsen, 1997 ; Paulsen et al. , 1998 ; Manfredi
et al. , 2001 ) and glenoid fossa remodelling ( Woodside et al. ,
1987 ; Ruf and Pancherz, 1999 ; Voudouris et al. , 2003a , b ).
The Herbst appliance was fi rst introduced in 1905 ( Herbst,
1934 ). Schwarz (1934) presented a number of improvements
of the dental anchor age system. After a period of virtual
oblivion, the Herbst appliance was re-introduced by Pancherz
(1979) . Since then, several modifi cations of the original
design have appeared. Examples of this are the bonded
Herbst appliance with upper and lower acrylic splints ( Howe,
1982 ), the cast splint Herbst appliance in combination with
headgear ( Wieslander, 1984 ), the Herbst appliance with a
mandibular acrylic splint and stainless steel crowns on the
maxillary fi rst molars ( Valant and Sinclair, 1989 ), and cast
splints equipped with a Herbst mechanism with ball and
socket hinges (Herbst IV®).
The integrated Herbst appliance — treatment effects in a group
of adolescent males with Class II malocclusions compared with
growth changes in an untreated control group
Paul Hägglund *, *** , Staffan Segerdal **, *** and Carl-Magnus Forsberg ****
* Department of Orthodontics, University of Umeå , ** Clinic of Orthodontics, Sundsvall , *** Department of Research
and Development Västernorrland County Council and **** Karolinska Institutet, Huddinge, Sweden
SUMMARY In this study, the effect of the integrated Herbst appliance (IHA) was examined in 30 Swedish
males (mean age 14.2 ± 0.96 years) with a Class II malocclusion. An evaluation of hand – wrist radiographs
showed that the patients were in the maturation stages MP3-F, MP3-FG, or MP3-G at the start of
treatment. The average treatment time with the Herbst mechanics was 0.7 years. Dentoskeletal and soft
tissue parameters were analysed on lateral radiographic head fi lms taken at the start and end of the IHA
treatment. The pre- and post- H erbst values of a number of skeletal and dental variables in the treatment
group were compared with the corresponding values in a group of untreated age-matched males with
Class II malocclusions. Differences in the cephalometric measurements pre- and post-Herbst treatment
were determined using paired t -tests.
In general, the control group exhibited only minor or no changes during the period of observation,
whereas treatment with the IHA resulted in statistically signifi cant and favourable changes of the
recorded variables. In the IHA patients, ANB angle was reduced on average by 2.1 degrees. However, a
skeletal post-normality (ANB = 3.9 degrees) remained even though a Class I dental relationship had been
obtained. In comparison with treatment effects achieved with other designs of Herbst appliances, some
minor differences in the changes of the variables SNA and ML/NSL were noted in the present study.
These differences could probably be attributed to the particular treatment protocol which was applied in
the IHA treatments.
The design of the Herbst appliance used in the present
study has been previously described in detail ( Haegglund
and Segerdal, 1997 ). The development of this appliance
was started in 1986 when the authors focused on fi nding a
simple way of integrating the Herbst mechanism with a
standard upper and lower fi xed appliance. With such an
arrangement, the complete fabrication of the Herbst
appliance could be performed in the clinic without
involvement of a dental technician. Furthermore, possible
problems with the appliance during treatment could be
immediately and easily solved in the clinic, making
involvement of a dental laboratory unnecessary.
The aim of the present investigation was to study the
treatment effects of this modifi ed Herbst appliance on a
number of dentofacial variables and compare the skeletal,
dental, and soft tissue profi le changes with the corresponding
morphological changes of an untreated group of subjects
with a post-normal occlusion.
Materials and methods
The records of 174 patients who had undergone treatment
with the integrated Herbst appliance (IHA) at the Clinic of
Orthodontics in Sundsvall, Sweden, were analysed as a part
by guest on June 2, 2013
THE INTEGRATED HERBST APPLIANCE —TREATMENT EFFECTS
of a treatment quality project. The following radiographic
records were available for each patient:
1. Standardized lateral cephalometric radiographs at the
start and termination of Herbst treatment.
2. Hand-wrist radiographs taken at the time of insertion of
the Herbst appliance.
From these patients, a consecutive group of Caucasian
boys who fulfi lled the criteria listed below was selected for
the present investigation.
1. No extractions of the permanent teeth either before or
during Herbst treatment.
2. The skeletal maturation phase at the start of Herbst
treatment should be MP3-F, MP3-FG, or MP3-G (from
onset of the pubertal growth spurt to peak) as evaluated
on the hand-wrist radiographs ( Hägg and Taranger, 1980 ,
1982 ). Two investigators (PH and SS) evaluated the
radiographs. Any disparities in the assessment of skeletal
maturation were discussed until consensus was reached.
3. ANB angle ≥ 4 degrees and overjet ≥ 6 mm after levelling
of the teeth with fi xed appliance.
Thirty-one boys fulfi lled these criteria. However, one boy
exhibited a considerable delay in skeletal maturation as a
consequence of which his chronological age at the start of
treatment was comparatively high. Since the control records
did not cover such high age levels, this subject was excluded
from the study.
The dental maturity of the patients corresponded to the
dental stages DS4/M1 ( n = 2) or DS4/M2 ( n = 28).
All patients had at least an end-to-end Class II molar
relationship before treatment. In three of the 30 patients, the
malocclusions were diagnosed as Class II division 2. During
the alignment phase, i.e. before insertion of the IHA, these
dentitions were changed to a Class II occlusion with
normally inclined or slightly proclined upper incisors.
Ten of the patients had previously undergone activator
treatment with poor results due to lack of compliance.
The Herbst treatments had been carried out by three
orthodontists using basically the same technique. At the
start of treatment, the mean age of the subjects was 14.2 ±
0.96 years (range 12.5 – 16.2 years). At this stage, three
patients were in maturation phase MP3-F, three in MP3-FG,
and 24 in MP3-G. The average treatment time with the IHA
was 8 months and 13 days (0.70 years, range 0.5 – 0.9 years).
Consequently, the mean post- H erbst age was 14.9 ± 0.95
years (range 13.2 – 16.8 years).
Before insertion of the IHA, the dental arches were
levelled with upper and lower fi xed appliances, which
included bands with headgear tubes on the upper fi rst molars
and bands with a rectangular and a round tube (0.036 – 0.045
inch) on the lower fi rst molars. The inclination of the upper
and lower anterior teeth was also adjusted, and the incisors
were intruded when necessary. After this initial treatment
procedure, which took on average 9.3 months, heavy
rectangular stainless steel arch wires in the upper and lower
jaw and the Herbst appliance could be inserted ( Haegglund
and Segerdal, 1997 , Figure 1 ). Apart from some minor
improvements of the connection between the Herbst
components and the fi xed appliance, no changes to the
appliances were carried out during the period of
The control subjects, who were of Austrian origin, were
selected from the cross-sectional roentgen cephalometric
material of orthodontically untreated children and
adolescents presented by Droschl (1984) . The diagnosis of
a Class II malocclusion was applied when there was a Class
II relationship exceeding one-quarter premolar width,
increased overjet, and incisor protrusion. In the age interval,
which corresponds to the pre- H erbst age range of the
treatment group (12.5 – 16.2 years), this material comprised
64 males with post-normal occlusion. Seventeen of these
subjects had to be excluded due to the fact that they exhibited
an ANB angle and/or an overjet smaller than 4 degrees and
6 mm, respectively. After this reduction, the control group
for pre- H erbst comparisons comprised 47 subjects. The
control group for the post- H erbst comparisons was selected
in a similar way, and 33 control subjects were available in
the age range 13.2 – 16.8 years.
According to the Greulich and Pyle (1959) method, the
skeletal age of the controls was estimated to be retarded, 4.6
months on average, in relation to chronological age.
However, when this assessment was performed on the same
material using the method of Tanner and Whitehouse
( Wenzel et al. , 1984 ), a closer agreement between skeletal
and chronological age was found. These results suggest that
the majority of the control subjects could be expected to be
in the maturation stages MP3-F, MP3-FG, or MP3-G ( Hägg
and Taranger, 1980 , 1982 ; Hägg and Pancherz, 1988 ).
Figure 1 The integrated Herbst appliance. The Herbst pistons are
connected to an auxiliary arch wire (0.9 – 1.0 mm). This arch wire is inserted
in round tubes on the lower fi rst molar bands and anteriorly attached to the
main arch wire with elastomeric modules. The Herbst tubes are attached to
the headgear tubes on the upper fi rst molar bands with hooks made of
0.9 – 1.0 mm hard stainless steel wire.
by guest on June 2, 2013
P. HÄGGLUND ET AL.
All radiographic registrations of the patients were made
with the same X -ray equipment and by the same operator.
The head fi lms were exposed with the patients standing
with the teeth in centric occlusion (intercuspal position:
ICP). The enlargement factor was 9.1 per cent. Corrections
for linear magnifi cation were not made.
A matte acetate tracing fi lm was placed on the lateral
head fi lm and relevant reference points ( Figure 2 ) and lines
( Figure 3 ) were marked with a pencil (0.3 mm 2H). All
tracings were digitized with a Scriptel RTD digitizer
(Scriptel Corporation, Toronto, Canada), and measurements
of the angular and linear variables listed in Table 1 were
performed with an accuracy of 0.1 mm or 0.1 degree using
the Dentofacial Planner computer program (Dentofacial
Planner Software Inc., Toronto, Canada).
The evaluation of treatment changes was carried out by
comparing the pre- and post-treatment tracings. This
procedure was based on superimposition of the tracings on
stable cranial base structures according to the technique
described by Björk and Skieller (1983) .
The corresponding individual measurements in the
control material were obtained directly from the author
( Droschl, 1984 ). The magnifi cation factor of the
cephalometric data of the control material was not
known. However, in order to be able to compare linear
measurements in patients and controls, it was necessary to
establish the magnifi cation factor in the control group. A
conceivable way of obtaining an acceptable estimation of
the magnifi cation factor would be to calculate it on the
basis of the length of the cranial base. Craniometric studies
of Caucasian, African, and Asian skulls of different ethnic
origins and shapes have shown that the variability of
Figure 2 Reference points used in the cephalometric analysis. The
defi nitions of the reference points are those given by Droschl (1984) .
Figure 3 Reference lines used in the cephalometric analysis. Defi nitions
of ML, NL, NSL, OL, RL, and Ils lines are those given by Björk (1960) .
The Apg and E-lines have been defi ned according to Ricketts (1957)
Table 1 The intercept values of the variables recorded in the
patients and controls at 14.2 years of age (pre-Herbst). The level of
statistical signifi cance of differences between the groups was
calculated with t -test.
VariablesIntercept values Difference
n = 30
n = 47
* P < 0.05, ** P < 0.01, *** P < 0.001.
by guest on June 2, 2013
THE INTEGRATED HERBST APPLIANCE —TREATMENT EFFECTS
cranial base length is small ( Martin and Saller, 1959 , Kuroe
et al. , 2004 ). Roentgen cephalometric studies of the length
of the anterior cranial base of individuals with different
ethnic backgrounds have also shown that the mean values
and standard deviations are comparable in the different
groups ( Solow and Sarnäs, 1982 ; Kerr and Ford, 1986 ,
Dibbets and Nolte 2002 ). To the best of our knowledge, a
comparison of anterior cranial base length between
Swedish and Austrian subjects has not been carried out.
However, there is no reason to believe that such a
comparison would produce a result which differs from
those of previous studies in this fi eld. When not corrected
for radiographic enlargement, the anterior cranial base
length at the start of Herbst treatment was on average 75.3
mm in the patient group. The corresponding measurement
in the control group was on average 75.0 mm. Post-Herbst
these distances were 76.9 and 76.5 mm, respectively,
in the treated and control groups. On the basis of these
observations, it was considered reasonable to assume that
the magnifi cation factors which apply to the two groups
were nearly identical.
Two-thirds of the patient cephalograms had been analysed
by one author (PH) and one-third by another author (SS)
Before the start of the tracing procedure, a calibration was
performed between the examiners as regards the identifi cation
of the landmarks. The concordance in landmark identifi cation
was tested by double determinations of the relevant angular
and linear measurements of 10 randomly chosen head fi lms.
The interindividual method errors ( s i ) were then calculated
according to Dahlberg’s (1940) formula:
where ‘ d ’ is the difference between the fi rst and second
measurements and ‘ n ’ is the number of double determinations.
The method errors of the angular measurements were found
to vary between 0.2 (ANB) and 0.9 (RL/ML) degrees, with
the exception of the variable Ils/NSL which exhibited an
error of 1.7 degrees. The method error of the linear
measurements varied between 0.5 and 0.6 mm.
The infl uence of the method error on the cephalometrically
determined treatment changes was evaluated using the pre-
and post- H erbst head fi lms of 20 randomly selected patients.
Tracings and measurements of both the pre- and the post-
H erbst fi lms of each patient were performed on two
occasions with an interval of at least 3 months. The post-
H erbst tracings obtained on the fi rst occasion were
superimposed on their respective pre- H erbst tracings and
the sizes of the changes (changes A) which had occurred
during treatment could be determined. Subsequently, the
corresponding changes (changes B) were determined on the
basis of the second set of pre- and post- H erbst tracings.
Finally, the differences between changes A and B were
entered into Dahlberg’s (1940) formula (see above) and a
combined method error, which included errors in the
location of landmarks as well as superimposition and
measurement errors, could be calculated for each variable.
The method error for linear variables varied between 0.5
(ii-Apg, ii-npg, is-npg, overbite, Li-E-line) and 0.8 mm (s-
gn). The corresponding values for angular measurements
were 0.5 (SNB, ANB, snpg) and 1.4 degrees (Ils/NL).
In the control material, the method error according to
Dahlberg’s (1940) formula (based on double determinations
carried out on 12 cephalograms) varied between 0.3 (is-
npg) and 1.2 (overbite) mm for linear measurements and
between 0.7 (snpg) and 1.5 (ML/NL) degrees for angular
measurements ( Droschl, 1984 ).
Within the treatment and control groups, a regression line
of the pre-Herbst cephalometric measurements on age was
calculated for each variable: y = a + b × age + c × group.
The observations for each variable were plotted in
diagrams in order to check that the linear regression
actually gave a reasonable fi t. Through geometric
translation of the y -axis, the intercepts ( a ) and the standard
errors of the intercepts were calculated ( Draper and Smith,
1966 ) at the age of 14.2 years, which was the mean age at
the start of treatment. Comparisons of the estimated
intercepts ( a ) for patients and controls were then performed
with the independent t -test. Corresponding regression
lines were formed for the post-Herbst cephalometric
measurements, and comparisons were then made of the
calculated intercepts for patients and controls at 14.9
years, which was the mean age at the end of the treatment
with the Herbst mechanics.
Differences between the
cephalometric measurements in the patient group were
tested for signifi cance with a paired t -test.
pre- and post-Herbst
The cephalometric variables in the Herbst and control
groups before treatment are compared in Table 1 . With the
exception of RL/ML, which was 3.5 degrees larger in the
control group ( P < 0.05), there were no signifi cant differences
between the skeletal variables of the patient and control
groups. However, the linear variables which describe the
position of the incisors differed, on average, between 1.3
and 1.9 mm between the groups. These differences could
partly be explained by the fact that alignment of teeth, and
in some cases also intrusion of the incisors, had been carried
out in the patients before insertion of the IHA.
Table 2 shows the changes which occurred in the patients
during treatment with the IHA. With the exception of the
variables NL/NSL, RL/ML, and Li-E-line, all linear and
by guest on June 2, 2013
P. HÄGGLUND ET AL.
angular measurements exhibited statistically signifi cant and
After an average treatment time of 8 months 13 days
(0.70 years) with the IHA, the ANB angle in the patient
group was signifi cantly smaller than that of the control
group ( P < 0.001; Table 3 ). In the vertical plane, the patients
exhibited a signifi cantly greater posterior face height (s-go)
than the controls ( P < 0.05) post- H erbst. As regards dental
measurements, there were six signifi cant differences
between the patients and controls. The distance between the
incisal edge of the upper incisors and the npg line was
reduced in the treatment group and was signifi cantly smaller
than in the controls ( P < 0.01). The incisal edge of the lower
incisors had been anteriorly displaced as a result of IHA
treatment and, in relation to the Apg line, this distance was
signifi cantly larger ( P < 0.05) in the patients as compared
with the controls.
The signifi cant changes in incisor positions in the patients
were also refl ected in the overjet value, which was normal
after treatment. In the control group, this value was 7.9 mm
and signifi cantly larger than in the treatment group ( P < 0.001).
The overbite had been reduced to 1.9 mm in the patients, while
no such change had occurred in the controls, and the mean
values in the two groups were signifi cantly different ( P <
Finally, normalization of the overjet in the patient group
resulted in an improved upper lip position in relation to
the E-line, and this variable value was signifi cantly smaller
in the patient group as compared with the controls
( P < 0.05).
The effects of 0.70 years of treatment with the IHA on
dental and skeletal variables in a group of male patients
with a post-normal malocclusion were evaluated in this
investigation. The study was performed on cephalograms
taken of clinical material which was well defi ned as regards
gender, skeletal maturation, diagnosis, and method of
treatment. It would have been desirable to have access to
control material of similar quality. However, for ethical
reasons it is no longer possible to perform longitudinal
roentgenologic registrations of untreated patients with
post-normal occlusion. As a compromise, therefore, it was
considered acceptable to use cross-sectional material
comprising age-matched untreated males with post-normal
occlusion ( Droschl, 1984 ) as a control group.
Information regarding the roentgenographic enlargement
was lacking for the control material. However, a comparison
of the average anterior cranial base length in the patient and
control groups showed that these values were nearly
identical. Studies have shown that cranial base length
exhibits very little variation between groups of different
ethnic origins ( Martin and Saller, 1959 ; Dibbets and Nolte,
2002 ; Kuroe et al. , 2004 ), and on this basis it is reasonable
to presuppose that the roentgenographic magnifi cation
factor was the same in the two materials.
Table 2 Mean changes ( d ¯ ) and standard deviations (SD) of the
cephalometric values after 0.7 years of treatment with the
integrated Herbst appliance in 30 boys.
* P < 0.05, *** P < 0.001.
Table 3 The intercept values of the variables recorded in the
patients and controls at 14.9 years of age (post-Herbst). The level
of statistical signifi cance of differences between the groups was
calculated with the t-test.
VariablesIntercept values Difference
n = 30
n = 33
Levels of signifi cance: * P < 0.05, ** P < 0.01, *** P < 0.001.
by guest on June 2, 2013