Download Analysis of anterior dentoalveolar and perioral aesthetic

European Journal of Orthodontics 36 (2014) 719–726
doi:10.1093/ejo/cjt102
Advance Access publication 7 February 2014
© The Author 2014. Published by Oxford University Press on behalf of the European Orthodontic Society.
All rights reserved. For permissions, please email: [email protected]
Analysis of anterior dentoalveolar and perioral aesthetic
characteristics and their impact on the decision to undergo a
Phase II orthodontic treatment
Carlos Flores-Mir*, Matthew M. Witt**, Giseon Heo*,*** and Paul W. Major*
*School of Dentistry, University of Alberta, Canada, **Private Practice, Delta, BC, Canada, ***Department of
Statistics, University of Alberta, Canada
Correspondence to: Carlos Flores-Mir, School of Dentistry, University of Alberta, 5–528 Edmonton Clinic Health
Academy, 11405–87 Ave NW, 5th Floor, Edmonton, Alberta, T6G 1C9, Canada. E-mail: [email protected]
summary Objectives: Researchers have conducted extensive studies regarding dentoalveolar factors that affect
anterior dental aesthetics; however, there is no consensus regarding how these factors affect orthodontic
treatment decisions. Only a few studies have included multiple factors simultaneously. Therefore, the
objective was to investigate if there are identifiable dentofacial and perioral aesthetic factors that bias
laypeople towards discontinuing treatment after a phase I treatment with this fixed class II corrector.
Methods: An analysis of photos and dental casts of 60 children (23 males, 37 females) having received
phase I orthodontic treatment with the Xbow appliance was conducted. Variables considered were incisor
height and width measurements, incisor proportions, incisor angulations, vertical lip thickness, gingival/
incisal display, smile width per cent, diastema, midline deviation, smile arc, gender, and use of a 2 × 4.
A principal component analysis and a logistic regression were used to determine which factors related to
a patient’s likelihood of receiving further orthodontic treatment.
Results: Only the angulation of the right maxillary incisors was significantly related to a patient’s likelihood
(odds ratio 1.886 (1.004–3.466); P = 0.049) to proceed to phase II orthodontic treatment following phase
I orthodontic treatment with the Xbow appliance. The odds of proceeding to phase II treatment were 86.6%
greater with a one standard deviation increase in the angulation of the right central and lateral incisors. Other
factors demonstrated trends but were not statistically significant.
Limitations: Sample in subgroups was small, excluded smiles that did not expose the upper incisor crowns
significantly, smiles in real life are observed three-dimensionally, other factors outside the aesthetic measurements were not considered in the analysis.
Conclusions: In this sample, the angulation of the maxillary right incisors was the most significant factor
influencing the decision to undergo an orthodontic phase II.
Introduction
Previous studies (Johnston et al., 1999; Moore et al., 2005;
Rosenstiel and Rashid, 2002; Parekh et al., 2006) have shown
that the lay public is able to identify factors that detract from
an aesthetic smile; however, they are far less critical than
dental professionals with regards to some of those elements
(Kokich et al., 1999, 2006; Martin et al., 2007). Tooth shape,
tooth size and proportion, incisor position (including tooth
angulation and presence of a diastema), midline deviation,
gingival display and morphology, smile arc, buccal corridors
are important factors for frontal dentofacial aesthetics from
the layperson perspective as shown by two related systematic
reviews (Witt and Flores-Mir, 2011a,b). Studies investigating these factors are typically survey based, with laypeople ranking or scoring a number of frontal photographs of
smiling patients. In the majority of cases, photographs are
either selected or modified such that only one parameter varies between the different photographs presented as survey
stimuli. Only a few studies (Hulsey, 1970; Parekh et al.,
2007; McNamara et al., 2008) have attempted to analyze
multiple factors simultaneously to determine their relative
influence on the layperson’s perception of dental aesthetics. Moreover, there is limited research investigating how
these various factors interact simultaneously and influence a
patient’s decisions regarding orthodontic treatment.
Phase I treatment with the Xbow may or may not be followed by a phase II treatment with conventional braces. The
Xbow is a class II fixed inter-arch corrector that supports
the Forsus springs on a custom-made metallic framework
attached to the upper and lower dentition (Flores-Mir et al.,
2009). It appears that a number of patients and parents are
satisfied with the occlusal and aesthetic results of the phase
I Xbow treatment and do not opt for phase II treatment
despite a recommendation to continue with treatment. This
recommendation is usually based on attaining final occlusal
and aesthetic details. It can be hypothesized that patients
720
that decide not to undergo a phase II orthodontic treatment
may be completely satisfied with the final attained occlusion even though the treating orthodontist suggest otherwise. The objective of this study was to investigate if there
are identifiable dentofacial and perioral aesthetic factors
that bias laypeople towards discontinuing treatment after a
phase I treatment with this fixed class II corrector.
Materials and methods
Ethics approval
Ethical approval was granted by the appropriate committee
at the University of Alberta.
Study sample
The treatment sample was obtained from the private practice and included all patients consecutively treated with the
Xbow appliance between early 2006 and late 2010 that had
dental casts, pre- (T1) and post-treatment (T2) digital frontal photographs of their smiles.
Photographs were taken at f-stop 9.5 and ISO 200 using a
Fuji Fine Pix S3 Pro (Fujifilm Canada, Mississauga, Ontario,
Canada) equipped with a Nikon 60 mm lens and SU-800
Nikon Wireless Speedlight Commander (Nikon Canada,
Mississauga, Ontario, Canada). Ideally, the patient photographs would be taken in maximum smile, but there was
no way to retroactively determine if the smiles captured in
the photographs were demonstrations of maximum smile, a
natural smile, or a forced smile. Patients were excluded from
the study if their smiling photographs depicted smiles that
did not display the majority of the maxillary incisors because
measurements of the teeth and gingiva would not be possible.
For each patient, it was noted if a 2 × 4 fixed appliance was
used during phase I, and if phase II treatment was administered, however, no information regarding age and phase
I treatment time was collected. The specific reason for not
continuing treatment was not recorded in the clinical chart.
Figure 1 Study sample details.
C. Flores-Mir et al.
A total of 158 patients (38% male versus 62% female) had
been treated with the Xbow appliance during the selected
time frame. Unfortunately, only 60 of the 158 patients from
this pool had the necessary T2 records available. The final
sample included 30 patients that did undergo phase II orthodontic treatment and 30 that did not. Details of the study
sample population are illustrated in Figure 1 (note that this
is a description of the sample, not an allocation tree).
Measurements
The width of the patients’ maxillary right central incisors were
measured on dental casts using a Max-Cal electronic digital
caliper (Fowler Canada, Brantford, Ontario, Canada) for the
calibration of measurements performed in the photographs.
Patient photographs were cropped vertically (between the columella and the superior portion of the mental protuberance) and
horizontally (between the inferior portions of the nasiolabial
sulci) without distorting the proportions of the teeth and lips.
Using Kodak Orthodontic Imaging Software (Carestream
Dental LLC, Atlanta, Georgia, USA), a number of easily
measured and previously reported aesthetically important
(Witt and Flores-Mir, 2011a,b) dental and peri-dental factors were measured from both the T1 and T2 photographs
of each patient. The width of each patient’s maxillary right
central incisor was input into the software to calibrate the
measurements. Measurements were recorded as a continuous variable in millimetres with two decimals.
Height and width measurement. Using the calibrated photographs, height (H) and width (W) measurements were taken
for the upper right and left central and lateral incisors (Figure 2).
These values are identified as ‘1.2 W’, ‘1.2 H’, ‘1.1 W’, ‘1.1 H’,
‘2.1 W’, ‘2.1 H’, ‘2.2 W’, and ‘2.2 H’.
Tooth proportions. Using the height and width (HW) measurements of the upper incisors (Figure 2), the width:height ratios
were determined for each incisor. These values are identified as
721
Aesthetics Impact on Orthodontic Treatment
‘1.2 HW’, ‘1.1 HW’, ‘2.1 HW’, and ‘2.2 HW’. Tooth width:width
proportions between the central and ipsilateral lateral incisors
were also determined. These values are identified as ‘1.2/1.1’
and ‘2.2/2.1’.
Vertical lip thickness. The vertical thicknesses of the
upper and lower lip were measured in the facial midline
(Figure 3). These values are identified as ‘U lip’ and ‘L lip’.
Gingival display or incisal coverage. The vertical height
of the gingiva between the zenith of the gingival margin of
the central incisors and the lower border of the upper lip was
measured and recorded as a positive value (Figure 4). If the
zenith of the gingival margin was covered by the lower border of the upper lip, an estimated negative value representing the distance covered was recorded. In cases where one
central incisor was covered and the other was exposed, the
positive value was recorded. Regardless of whether there
was positive or negative gingival display, the value is identified as ‘gingiva’.
Smile width per cent. The distance between the inner aspects
of the commissures of the lip was measured to give the intercommissural width (ICW) as measured from the inner aspect
of the buccal mucosa (Figure 5). The distance between the
Figure 2 Width and height measurements for tooth 1.1.
Figure 3 Vertical thickness measurements for the upper and lower lips.
buccal aspects of the most laterally situated teeth in the smile
was measured to give the smile width, identified as ‘SmW’.
The smile width per cent was obtained by dividing the SmW
by ICW. This value was identified as ‘Sm%’ and is a measure
of buccal corridor size.
Diastema. If present, the maxillary midline diastema was
measured (Figure 6). This value is identified as ‘diast’ and
was given a value of 0 if no diastema was present.
Midline deviation. The deviation of the midline was measured (Figure 7). The midline was determined by drawing a
vertical line through the middle of the tip of the nose and
through the middle of the philtrum of the lip. No indication
of the direction (right or left) was given. This value is identified as ‘MLDev’.
Incisor Angulations. The angulations of the four upper
incisors were measured with respect to the horizontal plane
of the image (Figure 8). These values are identified as ‘1.2
A’, ‘1.1 A’, ‘2.1 A’, and ‘2.2 A’.
Smile arc. As an approximate measure of the smile arc
(i.e. the degree to which the incisal edges of the maxillary
teeth follow the superior border of the lower lip), the vertical
Figure 4 Gingival display measurement.
Figure 5 Oral aperture width (top) and smile width (bottom)
measurements.
722
C. Flores-Mir et al.
Figure 6 Measurement of maxillary midline diastema.
Figure 8 Measurement of the angulation of the maxillary incisors.
Figure 7 Measurement of midline deviation.
Figure 9 Measurement of distances between maxillary incisal edges and
lower lip superior border.
distance between the incisal edges of the four upper incisors
were measured (Figure 9). These values are identified as ‘1.2
E-L’, ‘1.1 E-L’, ‘2.1 E-L’, and ‘2.2 E-L’. It is important to recognize that an increase in these values indicates a decrease
in the consonance of the smile arc with the contour of the
lower lip.
(PC) while retaining most of the information in the original variables. The logistic regression analysis included
the reduced variables as determined by the PCA with the
dependent dichotomous final outcome as started or not a
phase II orthodontic treatment.
All statistical tests were performed using the Statistical
Package for the Social Sciences 19.0 (International Business
Machines Corp, Armonk, New York, USA).
Oral hygiene and gingival inflammation. The level of oral
hygiene/gingival inflammation was also considered in the
analysis but was not included due to difficulty in assessing
this factor on smiling photographs. Tooth shape was also
initially considered; however, due to the degree of variation between patient photos, this factor could not be easily
assessed and discarded for the final analysis.
Statistical analysis
The analysis of the data involved three elements: 1. reliability analysis (intraclass correlation), 2. principal component
analysis (PCA), and 3. logistic regression.
Due to the large number of variables measured and limited sample size, it was not considered appropriate to carry
out a logistic regression analysis with all the variables
included. Therefore, the number of variables was reduced
using PCA, a statistical procedure that combines correlated
variables into a set of variables called principal components
Results
Reliability analysis
As part of a reliability analysis, each measurement was
repeated five times (at intervals of 1 week) by a one
researcher for 10 randomly selected patients. Reliability
testing suggested high intraclass correlation coefficients,
indicating strong intra-rater reliability and agreement for
1.2 H, U lip, L lip, gingiva, ICW, SmW, diast, 1.1 A, 2.1
A, and the E-L series of variables. The 1.2 W, 1.1 W, 1.1
H, 2.1 W, 2.1 H, 2.2 H, MLDev, 1.2 A, and 2.2 A variables
demonstrated adequate reliability, though the lower limits
for the confidence intervals on these variables were less
than desirable. The variable 2.2 W had poor reliability
(Table 1).
723
Aesthetics
Impact on Orthodontic Treatment
Table 1 Intra-rater reliability for variable measurements.
Variable
ICC* 95% confidence interval
1.2 W
1.2 H
1.1 W
1.1 H
2.1 W
2.1 H
2.2 W
2.2 H
U lip
L lip
Gingiva
Intercommissural width
SmW
Diast
MLDev
1.2 A
1.1 A
2.1 A
2.2 A
1.2 E-L
1.1 E-L
2.1 E-L
2.2 E-L
0.856 (0.683, 0.956)
0.879 (0.739, 0.963)
0.770 (0.554, 0.925)
0.844 (0.674, 0.951)
0.847 (0.668, 0.953)
0.838 (0.666, 0.949)
0.362 (0.110, 0.710)
0.756 (0.534, 0.919)
0.966 (0.909, 0.990)
0.959 (0.874, 0.989)
0.949 (0.881, 0.985)
0.965 (0.915, 0.990)
0.941 (0.863, 0.983)
0.982 (0.956, 0.995)
0.810 (0.618, 0.939)
0.821 (0.633, 0.943)
0.916 (0.793, 0.976)
0.893 (0.764, 0.968)
0.802 (0.598, 0.937)
0.985 (0.964, 0.996)
0.992 (0.980, 0.998)
0.980 (0.952, 0.994)
0.988 (0.971, 0.997)
Table 2 Correlation between PC and each variable based on
varimax rotation (with outliers).
Variable
PC1
PC2
PC3
PC4
PC5
1.2 E-L
1.1 E-L
2.1 E-L
2.2 E-L
1.2 WH
1.1 WH
2.1 WH
2.2 WH
1.2/1.1
2.2/2.1
1.2 A
1.1 A
U lip
L lip
Gingiva
Sm%
MLDev
2.2 A
2.1 A
0.926
0.934
0.942
0.867
0.253
0.208
0.008
−0.028
0.201
−0.064
−0.084
0.125
0.190
0.141
0.177
0.224
0.054
−0.018
0.053
0.109
−0.046
0.048
0.175
0.625
0.880
0.926
0.789
−0.046
−0.049
−0.110
0.044
−0.041
0.061
−0.004
0.203
0.060
−0.113
−0.092
−0.015
0.105
0.115
0.098
0.566
−0.188
−0.133
0.412
0.826
0.821
−0.048
0.249
−0.133
0.060
−0.225
0.556
−0.042
0.429
−0.077
0.005
0.127
0.074
−0.203
0.152
−0.110
−0.082
−0.013
0.110
0.165
0.843
0.814
0.117
0.055
0.349
−0.392
−0.274
0.508
0.183
−0.071
0.141
0.141
0.067
0.163
0.089
−0.088
−0.024
0.139
−0.091
−0.003
0.053
0.812
0.832
−0.622
0.010
0.204
−0.188
−0.033
Bolded numbers are those values were the ones selected for the
specific column (PC1 to PC5).
*ICC, intraclass correlation coefficient.
Principal component analysis
Two sets of data were produced: one in which extreme values (i.e. outliers) were removed and one in which they were
kept in the data set. Outliers were identified as starred values (*) on a boxplot graph designating a value further than
three interquartile ranges from the nearer edge of the box.
The PCA was conducted with and without these extreme
values to determine the best theoretically fitted set of new
variables to employ in a future logistic regression. Four
PC analyses procedures were performed in attempt to find
the best model: 1. no rotation PCA with outliers, 2. varimax rotation PCA with outliers, 3. no rotation PCA without outliers, and 4. varimax rotation PCA without outliers.
Of these four analyses, the varimax rotation PCA with the
outliers produced the model with the most coherent results
(Table 2). Due to the sample size limitations in this study,
only those loading values (i.e. correlation between PC and
each variable) that were greater than or equal to |0.600| were
interpreted as significant, since components loadings above
|0.600| are considered reliable regardless of a limited sample
size (Velicer and Fava, 1998). A total of five s (PC1–PC5)
were derived from the PCA. The remaining four variables
(Sm%, MLDev, 2.2 A, and 2.1 A) did not interact significantly between themselves. It has to be noted that what
would have been PC6 (2.2 A alone) was not included for
further analysis as 2.2 A had poor reliability as noted before.
The total variance explained by each PC (i.e. the degree
to which each component contributes to the description
of the aesthetic smile) is shown in Table 3. Note that each
Table 3 Total variance of smile explained (rotation sums of
squared loadings).
Component
Component
name
Eigenvalue
% of variance
Cumulative %
PC1
PC2
PC3
PC4
PC5
Smile arc
Intratooth
Intertooth
Right angle
Lips
4.461
2.812
2.530
1.930
1.637
22.303
14.058
12.652
9.652
8.187
22.303
36.361
49.013
58.665
66.852
subsequent factor is responsible for progressively less variation. Table 3 also lists the names given to the individual
components.
PC1 is dominated by four variables (1.2 E-L, 1.1 E-L, 2.1
E-L, and 2.2 E-L) and can be considered as the average of these
four variables that represent the vertical distance between the
incisal edges of the four upper incisors and the superior border
of the lower lip and thus was labelled ‘smile arc’.
PC2 is dominated by four variables (1.2 WH, 1.1 WH,
2.1 WH, and 2.2 WH) and can be considered as the average of these variables that represent the tooth width:height
ratios (i.e. the proportionality within each tooth). This PC
was labelled ‘intratooth’.
PC3 is dominated by two variables (1.2/1.1, 2.2/2.1) and
can be considered as the average of these variables representing the width:width ratios between the central and ipsilateral incisors (i.e. the proportionality between teeth). This
PC was labelled ‘intertooth’.
724
C. Flores-Mir et al.
PC4 is dominated by two variables (1.2 A and 1.1 A) and
can be considered as the average of these variables representing the angulation of the right maxillary central incisor
and the right maxillary lateral incisor with respect to a horizontal plane. This PC was labelled ‘right angle’.
PC5 is dominated by three variables (U lip, L lip, and
gingiva) and can be considered as the average of these variables representing the vertical dimension of the upper and
lower lips. Gingiva display reacts accordingly. This PC was
labelled ‘lips’.
Independent t-tests were used to examine whether significant differences existed in any of the PC existed between
1. boys and girls and 2. those with and without 2 × 4. Results
indicated that those with 2 × 4 were significantly lower in the
right angle score than those without 2 × 4, (t(51) = 2.572,
P = 0.013) (Table 4). No other significant differences for
2 × 4 or sex were revealed.
Logistic regression
First, an exploratory logistic regression was performed
with the gender, 2 × 4, and diastema variables as covariates in order to test the hypothesis that these three variables
are strong contributors to the decision to proceed to phase
II treatment. None of the factors were significant. A second exploratory logistic regression was run including the
gender, 2 × 4, and diastema variable as covariates with the
five PC. Only 2 × 4 and PC4 (right angle) were significant.
Finally, through a forward stepwise logistic regression, only
right angle barely reached significance (odds ratio 1.866
(1.004–3466); P = 0.049). Thus the odds of proceeding to
phase II treatment were 86.6% greater with a one standard
deviation increase in the value of right angle.
Discussion
The goal of this study was to determine which aesthetic factors were related to a patient’s likelihood to proceed with
phase II orthodontic treatment after receiving phase I Xbow
orthodontic treatment. In this situation, the decision to continue to phase II treatment was viewed as an indicator of
the aesthetic result following phase I treatment. Nine frontal
Table 4 Independent samples t-test of 2 × 4 and the PCs*
Component
T score
Degrees of
freedom
P value**
Mean
difference
Right angle
Smile arc
Intratooth
Intertooth
Lips
2.572
0.448
−0.038
−1.700
0.841
51
51
51
51
51
0.013
0.628
0.970
0.095
0.404
0.780
0.322
0.322
0.314
0.320
*Equal variances assumed.
**2 tailed.
dentofacial aesthetic factors were chosen based on previous
research (Witt and Flores-Mir, 2011a,b). The results of a
PCA and logistic regression indicate that increased angulations of the maxillary right lateral and central incisor are
related to increased odds (1.866) that a patient will proceed
to phase II treatment.
Previous studies have shown that laypeople are sensitive
to changes in the angulation of the anterior dentition (either
as an occlusal plane cant or as individual teeth; Kokich et al.,
1999; Thomas et al., 2003; Wolfart et al., 2004; Geron and
Atalia, 2005; Gul-e-Erum, 2008; Ker et al., 2008) However,
none of these studies have determined how noticeable
changes to the angulation of the anterior dentition are in the
presence of other aesthetic defects. It is likely that seemingly separate aesthetic elements can influence each other’s
contribution to smile aesthetics (Ker et al., 2008).
Angulations of the left maxillary central and lateral incisors did not appear to be a significant contributor to the
decision to proceed with phase II treatment. The variables
representing the angulations of tooth 2.1 and 2.2 did not
correlate well with any of the PC during dimension reduction (Table 3). Thus, they were not included in the logistic
regression. The poor correlations of 2.1 and 2.2 angulations
to the PC may be due to the lower reliability of these measurements (Table 1). To put this into a clinical perspective,
however, during the intra-rater reliability testing the largest discrepancy between measurements of the angulation of
tooth 2.1 was approximately 5 degrees (with the average discrepancy of approximately 3 degrees). The largest discrepancy between measurements of the angulation of tooth 2.2
was approximately 9 degrees (with the average discrepancy
of approximately 4 degrees). One study (Rodrigues et al.,
2009) has shown that laypeople are not sensitive to a 10
degree distal angulation of the lateral incisor. Thus, the discrepancy in measurements due to the decreased intra-rater
reliability is probably not of significance to the layperson.
A study (Geron and Atalia, 2005) demonstrated that laypeople are more sensitive to counter clockwise cants (versus clockwise cants), which in a sense result in the increase
in angulation of the maxillary right incisors (and concurrent decrease in angulation of the maxillary left incisors).
However, there is research that suggests that right-handed
people are more sensitive to details in the right side of an
asymmetric image (Mead and McLaughlin, 1992). A skeletal underlying cause should not be discarded in these scenarios. This would not explain the apparent increased sensitivity
to increases in angulation of the maxillary right incisors, but
may have been a factor during the measurement process and
perhaps could explain increased variability on the left incisors as mentioned above. Interestingly, handedness appears
to be neuropsychologically linked to right-versus-left aesthetic preferences, diminishing the effects of cultural cues
(Strachan, 2000). It is important to keep in mind that people
demonstrate an overall preference for symmetry over asymmetry (Hulsey, 1970; Strachan 2000; Wolfart et al., 2005).
725
Aesthetics Impact on Orthodontic Treatment
It is interesting to note that incisor angulation was a significant factor but the use of orthodontic brackets on the incisors (2 × 4) was not, as the two should be intimately related.
Sample size may have played a role in this regard. This indicates that although the two variables should be related, right
angle is a more precise variable than 2 × 4 when predicting
a patient’s likelihood of proceeding to phase II treatment.
There are only two previous studies (Hulsey, 1970;
McNamara et al., 2008) that are relatively similar in design
to this study. In all three studies, buccal corridors appear to
be of no significance. However, this study and the McNamara
study (McNamara et al., 2008) showed that smile arc was
of no significance, which is in contradiction to the study by
Hulsey (Hulsey, 1970). It is possible that these differences
are due to the use of adult subjects or the changes in what is
considered aesthetic over the decades since the 1970s.
Unlike this study, most research on frontal dentofacial
aesthetics assesses the importance of only one or two factors at a time (Witt and Flores-Mir, 2011a,b). The ability
of this study to analyze multiple aesthetic factors simultaneously is both an advantage and disadvantage. Discerning
the relative importance of multiple aesthetic factors is
valuable, but the large number of factors needed to fully
describe each photograph weakens the statistical analysis.
It is tempting to include every imaginable aesthetic parameter for fear of excluding an important one. Even if a large
number of factors were included, it is still possible that laypeople would be discriminating between photographs based
on some unaccounted parameter.
Limitations
It is possible that the sample used for this study was a source
of bias. The records used in this study came from a sample of 158 patients that was reduced due to a lack of T2
(post-phase I) records. Most of the patients that received
only phase I treatment had T2 records, but the majority of
the patients that received phase II treatment did not have T2
records. It is unclear why T2 records were taken for only
some of the patients receiving phase II treatment. It is possible that these cases represented exceptional cases that made
them desirable for documentation.
There are two possible sources of bias related to the
upper lip. Firstly, the sample of patients was further reduced
because patients were excluded from this study if their
smiling photographs depicted smiles that did not display
the majority of the maxillary incisors. In most cases, it
appears that poor display of the maxillary incisors was due
to a poorly posed smile rather than anatomical limitations
and could have been remedied by retaking the photograph.
Secondly, for those cases in which the posed smile was adequate but the maxillary incisors were partially covered by
the upper lip the degree of incisal coverage was estimated.
Therefore, the measurements for gingival display are more
accurate than those for incisal coverage. The degree of
estimation could have been reduced by taking incisogingival caliper measurements of the incisors on the patient models rather than just the mesiodistal measurements.
An outstanding issue is the fact that 2D static photographs
were considered when in real life the face is evaluated as
a 3D fully expressive structure. In addition, patient’s data
were not considered if their smiling photographs depicted
smiles that did not display the majority of the maxillary
incisors. This could in fact be a reason to be willing to continue orthodontic treatment for aesthetic reasons. The decision to consider only full smiling photos was to be able to
properly complete the required measurements.
Furthermore, there was a difference in the proportion of
males to females in the sample (38% male, 62% female)
versus the private practice from which the sample was taken
(42% male, 58% female). Previous research (Huang et al.,
2004) conducted in the neighbouring American state of
Washington shows that there is a sex distribution of orthodontic patients (36% male, 64% female) that is more reflective of the ratio found in this study population. Still, it is
unclear why the discrepancy exists between the gender ratio
of the study population and the private practice from which
it was obtained and thus indicates a possible source of bias.
For this study, the decision regarding phase II orthodontic
treatment is treated as a proxy for the patient’s (and likely their
family’s) feelings about the aesthetic result of phase I treatment. Perhaps, the most obvious shortcoming of this approach
is that it attempts to assign frontal dentofacial aesthetics as
the single reason for family’s orthodontic treatment decisions,
when these decisions are often made in a psycho-socio-economic context that includes the patient and caregiver burnout,
peer pressures, and financial pressures. It was not possible to
contact the patients in this study to find out the reasons they
did or did not continue on to phase II treatment; even if it were
possible, the actual truth may not be disclosed by the patient or
their family (recall bias or fear to answer the truth).
The individuals from whom the photos were taken were
actual patients. Theoretically the decision to undergo further
treatment was likely made by a combination of the actual
patients but also their families. Hence, we use the term
‘laypersons’ as a more all-encompassing term to mean the
non-professional aspect of the decision to undergo further
treatment.
A prospective study overcoming the above-stated limitations but using a similar statistical analysis should be conducted to confirm or not the current findings.
Conclusions
The angulations of the maxillary right central and lateral
incisors were the most significant factor related to a patient’s
likelihood of receiving phase II treatment following phase
I treatment with the Xbow appliance; a one standard unit
increase in angulation of these teeth increases the odds of
proceeding with phase II by 87%.
726
References
Flores-Mir C, Barnett G, Higgins D W, Heo G, Major P W 2009 Short-term
skeletal and dental effects of the Xbow appliance as measured on lateral cephalograms. American Journal of Orthodontics and Dentofacial
Orthopedics 136: 822–832
Geron S, Atalia W 2005 Influence of sex on the perception of oral and
smile esthetics with different gingival display and incisal plane inclination. The Angle Orthodontist 75: 778–784
Gul-e-Erum, F M 2008 Changes in smile parameters as perceived by orthodontists, dentists, artists, and laypeople. World Journal of Orthodontics
9: 132–140
Huang G J, Marston B E, del Aguila M A 2004 Orthodontic care in an
insured population in Washington: demographic factors. American
Journal of Orthodontics and Dentofacial Orthopedics 125: 741–746
Hulsey C M 1970 An esthetic evaluation of lip-teeth relationships present
in the smile. American Journal of Orthodontics 57: 132–144
Johnston C D, Burden D J, Stevenson M R 1999 The influence of dental to
facial midline discrepancies on dental attractiveness ratings. European
Journal of Orthodontics 21: 517–522
Ker A J, Chan R, Fields H W, Beck M, Rosenstiel S 2008 Esthetics and
smile characteristics from the layperson’s perspective: a computer-based
survey study. Journal of the American Dental Association (1939) 139:
1318–1327
Kokich V O Jr, Kiyak H A, Shapiro P A 1999 Comparing the perception
of dentists and lay people to altered dental esthetics. Journal of Esthetic
Dentistry 11: 311–324
Kokich V O, Kokich V G, Kiyak H A 2006 Perceptions of dental professionals and laypersons to altered dental esthetics: asymmetric and symmetric situations. American Journal of Orthodontics and Dentofacial
Orthopedics 130: 141–151
McNamara L, McNamara J A Jr, Ackerman M B, Baccetti T 2008 Hardand soft-tissue contributions to the esthetics of the posed smile in
growing patients seeking orthodontic treatment. American Journal of
Orthodontics and Dentofacial Orthopedics 133: 491–499
Martin A J, Buschang P H, Boley J C, Taylor R W, McKinney T W 2007
The impact of buccal corridors on smile attractiveness. European
Journal of Orthodontics 29: 530–537
C. Flores-Mir et al.
Mead A M, McLaughlin J P 1992 The roles of handedness and stimulus
asymmetry in aesthetic preference. Brain and Cognition 20: 300–307
Moore T, Southard K A, Casko J S, Qian F, Southard T E 2005 Buccal
corridors and smile esthetics. American Journal of Orthodontics and
Dentofacial Orthopedics 127: 208–13; quiz 261
Parekh S M, Fields H W, Beck M, Rosenstiel S 2006 Attractiveness of
variations in the smile arc and buccal corridor space as judged by orthodontists and laymen. The Angle Orthodontist 76: 557–563
Parekh S, Fields H W, Beck F M, Rosenstiel S F 2007 The acceptability
of variations in smile arc and buccal corridor space. Orthodontics &
Craniofacial Research 10: 15–21
Rodrigues C d e D, Magnani R, Machado M S, Oliveira O B 2009 The
perception of smile attractiveness. The Angle Orthodontist 79: 634–639
Rosenstiel S F, Rashid R G 2002 Public preferences for anterior tooth variations: a web-based study. Journal of Esthetic and Restorative Dentistry
14: 97–106
Strachan A L 2000 In the brain of the beholder: the neuropsychological
basis of aesthetic preferences. Harvard Brain 7. http://www.hcs.harvard.
edu/~hsmbb/BRAIN/vol7-spring2000/aesthetics.htm (30 January 2014,
date last accessed).
Thomas J L, Hayes C, Zawaideh S 2003 The effect of axial midline angulation on dental esthetics. The Angle orthodontist 73: 359–364
Velicer W F, Fava J L 1998 Effects of variable and subject sampling on factor pattern recovery. Psychological Methods 3: 231–251
Witt M, Flores-Mir C 2011a Laypeople’s preferences regarding frontal
dentofacial esthetics: tooth-related factors. Journal of the American
Dental Association (1939) 142: 635–645
Witt M, Flores-Mir C 2011b Laypeople’s preferences regarding frontal dentofacial esthetics: periodontal factors. Journal of the American
Dental Association (1939) 142: 925–937
Wolfart S, Brunzel S, Freitag S, Kern M 2004 Assessment of dental appearance following changes in incisor angulation. The International journal
of prosthodontics 17: 150–154
Wolfart S, Thormann H, Freitag S, Kern M 2005 Assessment of dental
appearance following changes in incisor proportions. European Journal
of Oral Sciences 113: 159–165