ORIGINAL ARTICLE C

Effectiveness of early treatment of Class II

malocclusion
Timothy T. Wheeler, DMD, PhD,a Susan P. McGorray, PhD,b Calogero Dolce, DDS, PhD,a Marie G. Taylor, RDH,a
and Gregory J. King, DMD, DMScc
Gainesville, Fla, and Seattle, Wash

The purpose of this study was to examine and report the effectiveness of early treatment with the
headgear/biteplate and the bionator in patients with Class II malocclusion regardless of the mechanism of
correction and to compare early-treatment results with changes over a similar time period in an observation
group. The role of factors such as compliance was examined to determine their contribution to effective
treatment. The experimental design was a prospective, longitudinal, randomized controlled trial. At the end of
the early-treatment period, all 3 groups differed significantly (overall, P = .001) in percentage of treatment goal
achieved, with median values of 83% for the bionator group, 100% for the headgear group, and 14% for the
observation group. In both treated and observation subjects, the percentages of goal achieved varied by initial
molar class severity (treated, P = .0205; observation, P = .0040) and race (treated, P = .0314; observation,
P = .0416). Significant correlations in the treated subjects were identified between percentage of goal
achieved and bone age (13 bones) (r = 0.16; P = .037), bone age (20 bones) (r = 0.16; P = .043), compliance
(r = 0.26; P = .0005), and initial overjet (r = –0.26; P = .0095). Significant correlations were not detected in the
observation group. Sex, treatment group, age, mandibular plane angle, pretreatment, and retention did not
significantly affect percentage of goal achieved among the treated and the observation subjects. Correlation
between normalized compliance scores and percentage of goal achieved was high for both bionator (r = 0.50)
and headgear subjects (r = 0.49) at the end of treatment. Multivariate analysis suggested that headgear may
be superior to bionator/biteplane in achieving a Class II correction during early treatment. (Am J Orthod
Dentofacial Orthop 2002;121:9-17)

W
e have previously reviewed many important A majority of authors concluded that activator therapy
issues concerning the 2 timing strategies for also inhibits maxillary AP displacement, but that the
treating Class II malocclusion.1 Although headgear is more effective.2,3,5,6,9,10,12,15,18 The conse-
growth modification for Class II correction can be quite quence of these appliances on mandibular AP displace-
effective in certain persons, much remains unknown ment is less clear, with the majority of studies suggest-
regarding the mechanisms involved in the success or the ing no effect.2-5,7,11,13,15,19 A few studies suggest that
failure of these treatment approaches. activator therapy results in favorable AP mandibular
There is good agreement that the headgear/biteplate displacement.6,8,9,12,20 Some suggest that headgear ther-
and the bionator bring about favorable anteroposterior apy actually inhibits such displacement.17,19
(AP) apical base, molar relationship, and overjet The impact on the teeth has been investigated less
changes.2-14 However, there is disagreement over the frequently. Activators appear to retract the upper
mechanism of AP changes with these appliances. The incisor2,4,6,7,11,13 and, perhaps, the upper molar,6,7,9,18
clearest consensus concerns the inhibitory effect of with little effect on the mandibular teeth. Headgear
headgear therapy on maxillary anterior displacement.14-18 seems to cause only distal maxillary molar move-
ment.17,21
This study was supported by the National Institute of Craniofacial Reasearch, We have recently reported results from a random-
NIH grant DE08715. ized controlled trial (RCT) of the AP skeletal and den-
aDepartment of Orthodontics, College of Dentistry, University of Florida.
bDepartment of Statistics, Division of Biostatistics, College of Liberal Arts and tal changes after early Class II treatment with bionators
Sciences, University of Florida. and headgear.22 We observed that the headgear/
cDepartment of Orthodontics, School of Dentistry, University of Washington.
biteplane and bionator both enhanced mandibular
Reprint requests to: Timothy T. Wheeler, Box 100444, JHMHC, Department of
Orthodontics, University of Florida College of Dentistry, Gainesville, FL growth without detectable relapse a year after the end of
32610-0444; e-mail, [email protected],edu. active treatment. The combined skeletal and dental
Submitted, July 2000; revised and accepted, June 2001. effect of these appliances was an average of 0.5-0.75
Copyright © 2002 by the American Association of Orthodontists.
0889-5406/2002/$35.00 + 0 8/1/120159 mm per year of Class II correction. These data provoke
doi:10.1067/mod.2002.120159 the question of the clinical significance of this amount
9
10 Wheeler et al American Journal of Orthodontics and Dentofacial Orthopedics
January 2002

Fig 1. Study stages and data collection points for phase 1. OR, orthodontic records; DC, data col-
lection.

of correction when 7 to 8 mm of correction may be nec- clinical examination, medical and dental histories; maxil-
essary. However, only AP changes have been examined lary and mandibular impressions; centric occlusion bite
from these data to date. It is very possible that vertical registrations; lateral cephalometric, panoramic, and hand-
changes also bring about the Class II correction. wrist radiographs; and facial and intraoral photographs.
Regardless of these reports, we know that these appli- A stratified block randomization procedure was
ances do work in our patients. used to assign a treatment protocol to each subject.
Practitioners often cite patient compliance as a crit- Each subject had an equal likelihood of assignment to
ical role in success of treatment. Compliance with head- the observation, the retention, or the no-retention group.
gear has been shown to be extremely variable.23 To In addition, for those assigned to either retention or no
date, no studies have documented what level of compli- retention, it was equally likely that they were assigned
ance is necessary with different appliances for success- to treatment with either a headgear/maxillary retainer
ful treatment. with biteplane or a bionator.
The purpose of the present study was to examine Strata were defined by severity of Class II mal-
and report the effectiveness of early treatment with the occlusion (mild, bilateral one-half cusp; moderate,
headgear/biteplate and the bionator in patients with 1 side three-fourths cusp; severe, 1 side full cusp),
Class II malocclusion regardless of the mechanism of need for preparatory treatment, mandibular plane
correction. The role of factors such as compliance was angle (SN/mandibular plane, < 30°, 30°-40°, > 40°),
examined to determine their contribution to effective race, and gender (when mandibular plane angle was
treatment. We also examined changes over a similar greater than 40°).
time in untreated observation subjects. After assignment to a treatment group and any
preparatory treatment or observation, there were 3 nec-
MATERIALS AND METHODS essary stages: Class II early treatment, retention/
Research design nonretention, and follow-up. Each subject passed
The experimental design was a prospective, longitu- through each stage with data collection occurring as
dinal, RCT of treatment of children with a Class II mal- indicated in Figure 1.
occlusion. The complete details of this trial have been The headgear group received a cervical (MPA ≤ 40°)
previously reported.1 Briefly, Class II subjects were or high-pull (MPA > 40°) headgear and a flat-plane
identified, notified of the study, and, along with a par- maxillary acrylic anterior biteplane with labial bow and
ent, invited to participate. Inclusion criteria included: molar circumferential wires designed to disclude the
bilateral greater than or equal to one–half-cusp Class II posterior teeth. The headgear facebow did not touch the
molars, or unilateral greater than one–half-cusp Class II maxillary incisors. The circumferential wires of the
molars, fully erupted permanent first molars, not more biteplane were loosely adapted and clipped over the
than 3 permanent canines or premolars, positive overjet buccal tubes for retention. Subjects were instructed to
and overbite, and good general and dental health. wear the acrylic retainer/biteplane full time, removing it
Although the number of teeth was dictated in the inclu- only for eating, brushing, and contact sports; they were
sion criteria, all subjects possessed their deciduous sec- instructed to wear the headgear at least 14 hours per
ond molars at the beginning of early treatment or obser- day. The headgear were adjusted at each appointment to
vation. deliver 16 ounces of force per side.
After obtaining informed consent, standardized rou- Bionator subjects received a bionator to maintain the
tine orthodontic records were collected. These included a bite, with occlusal stops for maxillary and mandibular
American Journal of Orthodontics and Dentofacial Orthopedics Wheeler et al 11
Volume 121, Number 1

teeth. Subjects were instructed to wear the bionator 22 number of quarter-cusp units of change required to
hours per day, removing it only for eating, brushing, achieve a bilateral Class I molar relationship. The time
and contact sports. points considered were end of treatment (or similar
During the retention phase, both the bionator and the length of observation) and end of phase 1. Treatment
headgear/biteplane retention subjects were instructed to group differences were examined with chi-square tests
wear the appliance every other night to bed, approxi- for categorical variables and Wilcoxon rank sum tests
mately 10 hours a day. Subjects in the retention groups for ordinal or continuous variables.
discontinued use of their appliances. Factors examined included sex, race, pretreatment,
At each visit, the attending orthodontist encouraged treatment group, initial mandibular plane angle, initial
the children in each treatment group to wear their appli- overjet, initial bone-age scores and bone age from hand-
ances as instructed. Subjects with poor or no coopera- wrist radiographs,24 age at start of phase 1, and compli-
tion in following wear instructions and those with poor ance. Separate analyses were conducted for the treated
progress were not dismissed but were followed and had subjects and the observation subjects. Because compli-
data collected at each time point. ance scores were based on responses to different ques-
To control for proficiency bias, each child’s clinic tions for the 2 treatment groups, standardized compli-
appointments were rotated among the 4 project ortho- ance scores ([subject score–treatment group mean]/
dontists. Children in appliances were scheduled every treatment group standard deviation) were calculated for
month; children without appliances were scheduled those in each group for use in analysis. Relationships
once every 3 months. between outcomes and covariates were examined with
All appliances were removed by the clinic’s dental chi-square and Fisher exact tests for categorical vari-
assistant (a dentist) at each data collection appointment ables, and t tests, analysis of variance (ANOVA), and
before the examiner obtained any data, including radi- correlation estimates for continuous variables. Logistic
ographs, and replaced as necessary afterward, so that regression was used to investigate multivariate relation-
the examiners would be blinded to treatment group and ships between covariates and success at the end of
treatment stage. phase 1. A forward selection procedure was used, with
Class II treatment/observation (stage 2) ended when level 0.10 to enter, to aid in model building. Hosmer-
2 orthodontists independently agreed within a 3-month Lemeshow tests, which compare observed treatment
interval that at least a bilateral Class I molar relation- success with predicted probability of success, were used
ship existed, or 2 years had elapsed from the start of to evaluate the fit of the final models. For all analyses,
treatment. Stage 3 (retention/nonretention) and stage 4 a P value of less than .05 was considered statistically
(follow-up) were fixed 6-month intervals. The end of significant.
phase 1 occurred upon completion of stage 4. At each
visit and data collection point, molar class relationships RESULTS
(right and left) were scored on a 0 to 11 quarter-cusp Of the 277 subjects who began this study, 248
scale, with 5 representing a Class I relationship. (89.5%) completed phase 1. Descriptive information for
A compliance score, based on the orthodontist’s them is presented in Table I. Bionator, headgear, and
evaluation, was generated for each patient from data observation groups were similar for baseline variables
collected at each appointment during the treatment that could influence phase 1 treatment outcome. The
phase. All patients were assessed for attitude and oral treatment groups did not differ significantly by sex,
hygiene. Patients in appliances were assessed for break- race, initial mandibular plane angle, or molar class
age, reported wear, visual wear of the appliance, and severity. Pretreatment was not performed on the obser-
mobile molars in the headgear group, and coffin wire vation subjects unless necessary; this accounts for the
groove and pterygoid response in the bionator group. differences observed for this characteristic. Initial over-
Each response indicating compliance received a score jet and age were similar for the 3 treatment groups.
of 1; each response indicating a lack of compliance Bone scores and ages based on bone scores did not dif-
received a score of 0. A mean compliance score was cal- fer significantly (ANOVA P values > .50 for all com-
culated from all responses during the treatment phase. parisons).
At the end of treatment or observation, success was
Statistical methods achieved by 44% of those treated with bionators, 62%
Effective treatment was examined by using 2 out- of those treated with headgear, and 8% of those under
come variables. Success was defined as a bilateral Class observation (overall, P = .001; all 3 groups differed sig-
I molar relationship at a specified time. Percentage of nificantly). At the end of phase 1, success percentages
treatment goal achieved was calculated based on the were 38%, 50%, and 13% for the bionator, the head-
12 Wheeler et al American Journal of Orthodontics and Dentofacial Orthopedics
January 2002

Table I. Baseline characteristics Table II. Logistic

regression model of phase 1 success
for treated subjects
Bionator Headgear Observation P value
95%
N 79 90 79 Variable Odds ratio estimate Confidence interval
Sex, % female 38.0 41.1 38.0 .888 C
Race, % nonwhite 11.4 7.8 7.6 .633 C Molar class severity
Pretreatment, % 50.6 51.1 12.7 .001 C (compared with severe)
required Moderate 2.33 (1.51, 3.60)
Initial molar class .997 C Mild 5.44 (2.28, 12.97)
severity Nonwhite 4.57 (1.26, 16.54)
% mild 29.1 31.1 29.1 Male 2.01 (0.94, 4.32)
% moderate 24.1 24.4 25.3 Pretreatment 2.58 (1.20, 5.56)
% severe 46.8 44.4 45.6 Headgear 2.10 (1.02, 4.31)
Mandibular plane .983 C Bone age (13 bones) 1.31 (1.04, 1.66)
angle Compliance score 1.43 (0.98, 2.09)
% <30 25.3 26.7 22.8 (normalized)
% 30-40 67.1 66.7 69.6
% >40 7.6 6.7 7.6
Overjet
Mean (SD) 5.54 (2.49) 5.12 (2.21) 5.65 (2.48) .321 A were 3.3 times more likely to achieve success than were
Age those in the severe group; those with mild severity were
Mean (SD) 9.67 (1.09) 9.67 (0.81) 9.55 (0.91) .622 A 10.0 times more likely to achieve success, compared
C, chi-square test; A, analysis of variance; SD, standard deviation. with the severe group.
The final model for phase 1 success for treated sub-
jects included 7 variables (overall, P = .0001). Odds
gear, and the observation groups, respectively (overall, ratio estimates and 95% confidence intervals for these
P = .001; the headgear and the bionator groups did not variables are listed in Table II. As in the bivariate analy-
differ significantly, P = .16). sis, those with milder molar class severity and the non-
Initial molar class severity was associated with white subjects were more likely to have successful treat-
phase 1 success for both the treated and the observation ment. Older bone age (based on 13 bones) and a higher
subjects. In the observation group, success was compliance score also increased the likelihood of suc-
achieved by 30% of those with mild initial molar class cess. In addition, male sex and pretreatment increased
severity, 10% of those with moderate initial molar class the probability of phase 1 success. Of the 102 male sub-
severity, and 3% of those with severe initial molar class jects, 46% were successfully treated, compared with
severity (P = .007). For those treated, phase 1 success 42% of the females. Of the 86 subjects who required
percentages were 65%, 49%, and 29% for mild, moder- pretreatment, 49% had successful phase 1 treatment
ate, and severe molar class groups, respectively (P = compared with 40% of those who did not need pretreat-
.001). Race affected treatment outcome (P = .039), with ment. With adjustments for these characteristics, success
success in 69% (11 of 16) of the nonwhite subjects with headgear treatment was approximately twice as
compared with 42% (64 of 153) of the white subjects. likely, compared with bionator treatment. Lack of fit of
For treated subjects, those with successful phase 1 the model was not detected (P = .80).
treatment had less initial overjet (mean, 4.8 mm) than Similar patterns were observed for percentages of
did those with unsuccessful treatment (mean, 5.7 mm; molar class goal achieved (Fig 2). The mean goal for
P = .001). Compliance was significantly higher for each group was slightly less than 6 one–fourth-cusp units
those with successful treatment (mean compliance or approximately three-fourths cusp Class II on each
score, 0.22) compared with unsuccessfully treated sub- side. For percentages of goal achieved after treatment,
jects (mean, –0.17; P = .012). Sex, race, pretreatment, median values were 83% for the bionator group, 100%
mandibular plane angle, retention, age, and bone age for the headgear group, and 14% for the observation
did not significantly affect phase 1 success in either the group. All 3 groups differed significantly at the end of
treated or the observation subjects. treatment (overall, P = .001). Median values for percent-
Logistic regression, modeling phase 1 success as a age of goal achieved at the end of phase 1 were 75% for
function of covariates, was performed with separate the bionator group, 94% for the headgear group, and 17%
models developed for the observation and the treated for observation group. At the end of phase 1, percentages
subjects. For the observation group, only initial molar of goal achieved did not differ significantly between the
class severity entered the model (overall model, P = bionator and the headgear groups (P = .468), but all other
.0023). Those with moderate initial molar class severity comparisons were significant (P = .001 for all).
American Journal of Orthodontics and Dentofacial Orthopedics Wheeler et al 13
Volume 121, Number 1

Fig 2. Boxplots of percentage of treatment goal achieved at end of treatment and end of phase 1 by
treatment group. Box represents 25th, 50th (black bar), and 75th percentiles; whiskers show
extremes.

For both the treated and the observation subjects, relapsed. Relapse occurred more frequently in the head-
percentages of goal achieved varied by initial molar gear subjects (46%) than in the bionator (28%) and the
class severity (treated, P = .0205; observation, P = observation (19%) subjects. The opposite trend is seen
.0040) and race (treated, P = .0314; observation, P = for continued improvement; improvement was observed
.0416). Distributions of percentages of goal achieved in 48%, 30%, and 16% of the observation, bionator, and
for these characteristics are displayed by boxplots (Figs headgear subjects, respectively (overall, P = .001). For
3 and 4). Significant correlations of small magnitude treated subjects, 42% of those without retention had
were identified in the treated subjects between percent- relapsed compared with 32% of those with retention
age of goal achieved and bone age (13 bones) (r = 0.16; (P = .24). The average relapse in those without retention
P = .037), bone age (20 bones) (r = 0.16; P = .043), was 2.38 one–fourth-cusp units compared with 1.85
compliance (r = 0.26; P = .0005), and initial overjet units in those who underwent 6 months of retention
(r = –0.26; P = .0095). No significant correlations were (P = .09). Subjects who experienced relapse were more
detected in the observation group. Sex, treatment group, likely to be female (P = .022). They also had higher nor-
age, mandibular plane angle, pretreatment, and reten- malized compliance scores (P = .001) and were more
tion did not significantly affect percentage of goal likely to have successful treatment (P = .001). A logis-
achieved among the treated or the observation subjects. tic regression model for relapse identified 4 significant
The role of compliance in treatment result can best variables: treatment success (odds ratio estimate [OR],
be seen at the end of treatment. Correlation between 5.99), white race (OR, 4.72), treatment goal (OR, 1.37),
normalized compliance scores and percentage of goal and female (OR, 2.37). Once these variables are
achieved was high for both bionator (r = 0.50) and accounted for, the inclusion of compliance scores or
headgear subjects (r = 0.49) (Fig 5). treatment group did not significantly improve the
Molar class relapse, measured in one–fourth-cusp model. However, treatment group and compliance did
units, from the end of treatment to the end of phase 1 is affect treatment success, which was included in this
illustrated in Figure 6. In these plots, subjects located on model.
the diagonal had no change in percentage of goal
achieved between the end of treatment and the end of DISCUSSION
phase 1. Those located above the diagonal had contin- As reviewed above, both bionators and headgear
ued improvement, while those below the diagonal affect the craniofacial and dentoalveolar complex. We
14 Wheeler et al American Journal of Orthodontics and Dentofacial Orthopedics
January 2002

Fig 3. Boxplots of percentage of treatment goal achieved at end of phase 1 by treatment status and
initial molar class severity.

Fig 4. Boxplots of percentage of treatment goal achieved at end of phase 1 by treatment status and
race.

have previously shown that analysis of the AP data both headgear and bionator subjects, were stable a year
revealed that both bionator and headgear treatments after the end of treatment, but dental movements
corrected Class II molar relationships, reduced overjets relapsed.22
and apical base discrepancies, and caused posterior Jakobsson15 randomly assigned a group of 60 Class
maxillary tooth movement.22 The skeletal changes, II mixed-dentition subjects (mean age, 8.5 years) to
largely attributable to enhanced mandibular growth in observation, Andresen activator, and Kloehn headgear
American Journal of Orthodontics and Dentofacial Orthopedics Wheeler et al 15
Volume 121, Number 1

groups. He reported on treatment effects after 18

months but did not examine relapse. He reported that
both appliances reduced overjet in those groups com-
pared with the observation group, with a greater effect
in the activator group. Both appliances restricted maxil-
lary anterior displacement and had no effect on anterior
mandibular growth compared with observation; the
headgear was more effective in restraining maxillary
advancement than was the bionator. He did not report
individual tooth movements or on the effect of age,
severity, or gender.
Tulloch et al14,20 randomly assigned 175 Class II
subjects who had at least 7 mm of overjet to observa-
tion, bionator, and headgear groups, and treated or
observed them for 15 months. They reported that both
appliances reduced apical base discrepancies (ANB)
equally when compared with the observation group;
however, the mechanism of correction was different
between the 2 treated groups. The headgear restricted
maxillary anterior growth, whereas the bionator pro-
moted increased anterior mandibular growth. Their Fig 5. Plot of normalized compliance score vs percent-
study design did not include a retention or follow-up age of goal achieved at end of treatment with treatment
group indicated.
period but proceeded at 15 months to full appliance
treatment in all groups.
Although the changes observed in the above studies
have been shown to be statistically significant, the ques- uted the relapse of the treatment groups to dental and
tion arises as to the clinical significance due to the rela- not skeletal relapse.22 We interpreted this relapse to
tively small magnitude of the changes. This study used indicate that our retention protocol was only minimally
the power of the RCT to assess the treatment success of successful at maintaining dental correction, because
these appliances by examining the molar correction. 32% of those in the retention group had relapse (mean,
Clearly, early treatment with either bionators or 1.85 units) compared with relapse for 42% of those not
headgear/biteplane increased the number of patients retained (mean, 2.38 units). The role of compliance
achieving a Class I molar relationship over those not with retention protocol should also be examined. We
receiving early treatment. Approximately 44% of the noted that the headgear group showed a significantly
bionator and 62% of the headgear/biteplane patients greater increase in Class II correction during the treat-
achieved a Class I molar relationship during the treat- ment period. This correction relapsed during the obser-
ment phase, but only about 8% of the observation chil- vation period, equating the headgear to the bionator
dren self-corrected. Although these values represented a treatment effect. This is most likely due to the dental
complete correction to a Class I molar relationship, one relapse that occurred during this time. The percentage
may consider a Phase I case successful if the molar cor- of children with full-cusp Class II corrected to Class I
rects from a full-cusp Class II to a one–fourth-cusp during the treatment period was about 55%. This
Class II. This measure of success was depicted as the dropped to 30% during the retention observation year.
percentage of the goal achieved (Fig 2). The median One would hope that if the retention scheme could be
values for percentage of goal achieved ranged from improved to better retain the dental correction, this suc-
75% to 100% in the treatment groups and were about cess rate could be maintained.
15% in the observation group. This again demonstrates The severity of the Class II malocclusion was the
the effectiveness of the treatment methods in correcting only variable examined that significantly influenced the
a Class II molar relationship. outcome of phase 1 for both treated and observation
Both treatment groups demonstrated some relapse subjects. Approximately 30% of the mild, bilateral
of the Class II correction during the 1-year retention/ one–half-cusp Class II subjects in the observation group
observation period posttreatment, while the percentage self-corrected. Although this is a significant number,
of observation subjects achieving a Class II molar rela- treatment with either bionator or headgear increased
tionship increased to 13%. We have previously attrib- this number to 57% and 70%, respectively. The real
16 Wheeler et al American Journal of Orthodontics and Dentofacial Orthopedics
January 2002

Fig 6. Plots by treatment group of percentage of goal achieved at end of treatment vs end of phase
1 with compliance indicated. Points below diagonal indicate relapse.

value of this amount of correction in the children could sons. The nonwhite group was a small, heterogeneous
not really be assessed until the overall treatment was group, primarily blacks and children of Asian descent.
completed. This group also experienced a disproportionate amount
Permanent first-molar correction could be impacted of dropout; thus, bias might result. Further studies are
by the loss of the mandibular deciduous second molars. needed to clarify whether race plays an important role
Patients in 1 group might shed their deciduous molars at in treatment success.
different rates than do those in other groups; this could Compliance with treatment was difficult to accu-
then affect the outcome we examined. However, exam- rately assess and quantify. Although no 1 measure was
ination of the data at all the data collection points reliable, the aggregate of the measures we used showed
showed no significant differences between the groups in a correlation of increased success with increased com-
the number of patients with either 1, 2, or no mandibu- pliance. We do not know how much compliance is nec-
lar deciduous second molars present. essary, but it appears that orthodontists’ typical mea-
Modeling of treatment success suggests that many sures of compliance can predict successful molar
factors play roles. The impacts of sex, bone age, and correction.
compliance are difficult to interpret. Girls tended to
have higher compliance scores, particularly in the head- CONCLUSIONS
gear group, but boys tended to have younger bone ages. In this study, we have demonstrated that both
For both boys and girls, a trend of increasing success headgear and bionator treatments are effective in
with increasing bone age was observed. The regression achieving phase 1 treatment goals. Regression model-
analysis revealed that once other variables were ing results indicated that factors such as sex, race, ini-
accounted for, the treatment group had a significant tial molar class severity, and compliance also play
effect, with headgear treatment the more successful. roles, and, when these factors are taken into account,
This was not apparent in the bivariate results. Modeling headgear treatment is superior to bionator treatment.
of relapse suggested that those with the greatest This significant difference was identified, even
improvements (treatment success and large treatment though subjects treated with headgear experienced
goal) had an increased risk for relapse. The role of race more relapse between the end of treatment and the
in both models is difficult to interpret for several rea- end of phase 1.
American Journal of Orthodontics and Dentofacial Orthopedics Wheeler et al 17
Volume 121, Number 1

We thank the many graduate students, fellows, and 10. Vargervik K, Harvold E. Response to activator treatment in Class
dental students who helped with this study over the II malocclusions. Am J Orthod 1985;88:242-51.
11. Looi L, Mills J. The effect of two contrasting forms of orthodon-
years. We also want to express our most sincere thanks
tic treatment on the facial profile. Am J Orthod 1986;89:507-17.
to all the children and their families who participated in 12. Jakobsson S, Paulin G. The influence of activator treatment on
our study. Finally, a special heartfelt acknowledgement skeletal growth in Angle Class II:1 cases: a roentgenocephalo-
is given to Dr Stephen Keeling, who passed away in metric study. Eur J Orthod 1990;12:174-84.
January 1997. This study is a tribute to his dedication 13. Nelson C, Harkness M, Herbison P. Mandibular changes during
functional appliance treatment. Am J Orthod 1993;104:153-61.
for the advancement of clinical knowledge in ortho-
14. Tulloch J, Phillips C, Koch G, Proffit W. The effect of early inter-
dontics. vention on skeletal pattern in Class II malocclusion: a random-
ized clinical trial. Am J Orthod Dentofacial Orthop 1997;111:
REFERENCES 391-400.
1. Keeling S, King G, Wheeler T, McGorray S. Timing of Class II 15. Jakobsson S. Cephalometric evaluation of treatment effect on
Treatment: rationale, methods, and early results of an ongoing Class II, division 1 malocclusions. Am J Orthod 1967;53:446-56.
randomized clinical trial. In: McNamara J, editor. Orthodontic 16. Wieslander L, Tandlakare L. The effect of treatment on the con-
treatment: outcomes and effectiveness, monograph no. 30. Ann current development of the craniofacial complex. Am J Orthod
Arbor: Center for Human Growth and Development; University 1963;49:15-27.
of Michigan; 1995. p. 81-112. 17. Mills JRE. The long-term results of the proclination of the lower
2. Trayfoot J, Richardson A. Angle Class II, division 1 malocclusions incisors. Br Dent J 1966;120:355-63.
treated by the Andresen method. Br Dent J 1968;124:516-9. 18. Baumrind S, Korn EL, Isaacson RJ, West EE, Molthen R. Quan-
3. Harvold E, Vargervik K. Morphologic response to activator treat- titative analysis of the orthodontic and orthopedic effects of max-
ment. Am J Orthod 1971;60:478-90. illary traction. Am J Orthod 1983;84:384-98.
4. Wieslander L, Lagerstrom L. The effect of activator treatment on 19. Meach CL. A cephalometric comparison of bony profile changes
Class II malocclusions. Am J Orthod 1979;75:20-6. in Class II, division 1 patients treated with extraoral force and
5. Forsberg C, Odenrick L. Skeletal and soft tissue response to acti- functional jaw orthopedics. Am J Orthod 1966;52:353-70.
vator treatment. Eur J Orthod 1981;3:247-53. 20. Tulloch J, Proffit W, Phillips C. Influences on the outcome of
6. Luder H. Effects of activator treatment: evidence for the occur- early treatment for Class II malocclusion. Am J Orthod Dentofa-
rence of two different types of reaction. Eur J Orthod 1981;3: cial Orthop 1997;111:533-42.
205-22. 21. Wieslander L. The effect of force on craniofacial development.
7. Calvert F. An assessment of Andresen therapy on Class II division Am J Orthod 1974;65:531-8.
1 malocclusion. Br J Orthod 1982;9:149-53. 22. Keeling SD, Wheeler TT, King GJ, Garvan CW, Cohen DA,
8. Baumrind S, Korn E, Isaacson R, West E, Molthen R. Superim- Cabassa S, et al. Anteroposterior skeletal and dental changes fol-
positional assessment of treatment associated changes in the tem- lowing early Class II treatment with bionators and headgear. Am
poromandibular joint and mandibular symphysis. Am J Orthod J Orthod Dentofacial Orthop 1998;113:40-50.
1983;84:443-65. 23. Guray E, Orhan M. Selcuk type headgear timer (STHT). Am J
9. Johnston L. A comparative analysis of Class II treatments. In: Vig Orthod Dentofacial Orthop 1997;111:87-92.
PS, Ribbens KA, editors. Clinical judgment in orthodontics, 24. Tanner JM, Whitehouse NC, Marshall WA, Healy MJR, Gold-
monograph no. 19. Ann Arbor: Center for Human Growth and stein H. Assessment of skeletal maturity and prediction of adult
Development: University of Michigan; 1986. height (TW2 Method). 2nd ed. London: Academic Press; 1983.

AAO MEETING CALENDAR

2002 — Philadelphia, Pa, May 3 to 7, Pennsylvania Convention Center
2003 — Hawaiian Islands, May 2 to 9, Hawaii Convention Center
2004 — Orlando, Fla, May 1 to 5, Orange County Convention Center
2005 — San Francisco, Calif, May 21 to 26, Moscone Convention Center
2006 — New Orleans, La, April 29 to May 3, Ernest N. Morial Convention Center
2007 — Honolulu, Hawaii, April 27 to May 1, Hawaii Convention Center
2008 — Denver, Colo, May 16 to 20, Colorado Convention Center