Studies of differences in esthetic perception also have compared appraisals by dentists, parents and children, and have focused primarily on malocclusion and dental fluorosis. Pietilä and Pietilä⁸ found that parents’ concerns about their children’s dentition coincided with an orthodontist’s diagnosis of malocclusion 60 percent of the time, with the strongest agreement occurring when malocclusions affected the anterior teeth. Birkeland and colleagues⁹ found that children 11 and 15 years of age were more satisfied with their appearance than were their parents. Beyer and Lindauer⁴ concluded that orthodontists were less tolerant of midline deviations than were children and parents, and that parents were less tolerant than the children.

Evans and Shaw¹⁰ found that orthodontists agreed more frequently with parents than with children on the esthetics of a child’s anterior teeth. Using computer-generated images, McKnight and colleagues¹¹ had entering dental students rate various oral conditions. They found that the students viewed images of mild fluorosis less favorably than control images, images of mid-line diastemata less favorably than images of mild fluorosis and images of mild fluorosis less favorably than images of an isolated enamel opacity.

Fluorosis indexes do not focus on the child’s or the parent’s perception of the tooth’s esthetics. This omission is unfortunate because esthetics is the only significant consequence of dental fluorosis—at least within the levels of fluorosis commonly found in North America.¹² Researchers often have assumed that the primary measure of significance was the presence or absence of any dental fluorosis. Although certain discrete levels are detectable to dentists,¹³ fluorosis itself does not necessarily translate into dissatisfaction with the appearance of the teeth, for either the professional¹⁴ or the layperson. Only in the last decade has the issue been expanded beyond normative assessments.^11,15–26

Recognizing this paradox, researchers have started to examine patients’ and their parents’ perceptions of fluorosis. The psychosocial consequences of dental staining could be considerable.^19,27–31 Ellwood and O’Mullane¹⁷ found that the size and the severity of enamel opacities were related to the degree of satisfaction with the appearance of the anterior maxillary teeth. Clark²⁰ found that parents, children and dentists were able to distinguish between teeth with and without fluorosis. However, dentists rated the teeth’s esthetics significantly better than did parents for low levels of fluorosis or for nonfluorosis problems, and parents rated the esthetics significantly better than did the children. Lalumandier and Rozier²¹ found that parents whose children had a tooth surface index fluorosis, or TSIF, score greater than 0 had more than twice the odds of reporting dissatisfaction with their children’s dental esthetics than did the parents of children with a TSIF score of 0. Other studies have demonstrated that professionals and laypeople can distinguish between teeth that have fluorosis and those that do not, even if fluorosis is mild.^16,22,32 Clark and colleagues²³ found that parents were more critical of their children’s tooth color than were the children. Hawley and colleagues²⁴ showed 14-year-old children photographs of teeth corresponding to Thylstrup Fejerskov index, or TFI, scores of 0 to 4. Photographs corresponding to TFI scores of 2 and 3 were found to be unacceptable by 15 and 92 percent of the subjects, respectively. Stephen and colleagues²⁵ found that untrained lay observers unanimously rated almost four times more "homologous pairs of teeth with aesthetically unacceptable diffuse mottling." Table 1 summarizes the differences among dentists, parents and children found in some of these studies.^{3–5,8–10,20,23,26,33}

To better understand the esthetic perceptions of dentists, parents and children in North America, we conducted a study of how they evaluated the anterior teeth of children who were participating in the British Columbia Fluoridation Cessation Study. We used classification schemes that relied on the combination of the parents’, subjects’ and examiners’ opinions of esthetic issues; these schemes are supported by published research.²⁰

	METHODS AND SUBJECTS

TOP ABSTRACT METHODS AND SUBJECTS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Study population and sites. We used epidemiologic data and fluoride and diet histories from the follow-up surveys and questionnaires described in published studies.^18,26,34,35 Four standardized examiners assessed 8,281 children, who were in grades two, three, eight and nine during the 1996–1997 school year, for dental fluorosis using the TFI, in which scores range from 0 (no fluorosis) to 9 (severe loss of enamel with change of anatomical appearance).³⁶ Scores of 1 through 4 reflect increasing involvement of tooth surface with opaque areas and loss of translucency. Intra- and interexaminer agreement levels for dental fluorosis (on a per-child basis) were determined as a part of an intensive training program within the British Columbia Fluoridation Cessation Study. This study is a longitudinal evaluation of the effects of discontinuing the addition of fluoride to water supplies that had been fluoridated in British Columbia, Canada. The Comox/Courtenay and Campbell River communities stopped fluoridating their water supplies by 1992 (fluoridation-ended, or F-E, communities); the Kamloops community did so in 2001, so it is called the still-fluoridated, or S-F, community because the water was fluoridated when data were collected in the 1996–1997 school year. All of the examined subjects had their permanent anterior teeth mineralized at the time water supplies were optimally fluoridated in the three communities.

We hypothesized that girls have higher expectations about their tooth color than do boys, and parents of girls have higher expectations about their child’s tooth color than do parents of boys.

Examination conditions and procedure. These subjects were examined in nursing stations at schools using a portable dental chair and light. The examiners determined the TFI score after drying the subjects’ teeth with gauze. All teeth were viewed both with and without the use of the dental light to see contrasts and variations in different surfaces. The highest TFI score observed was assigned to each tooth. Examiners were blinded to data relating to a parent’s esthetic rating of their child’s tooth color and fluoride history. The examiners followed the Centers for Disease Control and Prevention’s standard procedures for epidemiologic dental examinations.³⁷ Informed consent was obtained from parents, as approved by the ethical review board of the University of British Columbia.

Reference statement. The dentist examiners’ (raters’), parents’ and subjects’ perceptions of the esthetic appearance of the anterior teeth were established by a single response to the closed-ended reference statement, "The color of these teeth (mine or my child’s) is pleasing and looks nice." They indicated their agreement or disagreement with the reference statement using a five-level Likert-type scale, on which a score of 1 = "strongly agree," a score of 2 = "agree," a score of 3 = "neutral," a score of 4 = "disagree" and a score of 5 = "strongly disagree." The reference statement conveyed the results of various lines of inquiry developed from past research, synthesizing the complex interactions into one question.²⁰

Structure of hypothesized contrasts. On the basis of a literature review and our collective clinical experience, we made the following hypotheses:

– girls have higher expectations about their tooth color (that is, they are more inclined to disagree with the reference statement) than do boys, and they have higher mean scores on the reference statement scale;

– parents of girls have higher expectations about their child’s tooth color than do parents of boys;

– older subjects have higher expectations about their tooth color than do younger subjects;

– parents have higher expectations about older children’s tooth color than about that of younger children;

– dentists have higher expectations about older children’s tooth color than about that of younger children;

– dentists have higher expectations about girls’ tooth color than about that of boys;

– parents with high socioeconomic status, or SES, have higher expectations than do parents with lower SES.

Statistical analyses. When several measurements are taken on the same subject, the measurements are correlated with one another. When the measurements could be thought of as responses to levels of an experimental factor of interest such as time, treatment or dose, the correlation could be taken into account by performing a multivariate, repeated-measures analysis of variance, or ANOVA. This commonly is done when measurements of a characteristic are taken on more than one occasion. Because we had assessments from three examiners (those for the dentist, parent and subject), we compared the three ratings using repeated-measures ANOVA using a statistical software program (Statistical Analysis System, Version 9.0, SAS Institute, Carey, N.C.). This is similar conceptually to a paired t test, but it allowed for the inclusion of covariates that would modify the raters’ scores and enabled us to ascertain differences in the perceived esthetic appearance of subjects’ teeth among three groups—parents, dentists and subjects.³⁸

We used the following variables in this analysis: sex, rater (dentist who examined the subject), community of residence, reason subject was dissatisfied with tooth color, SES (high or low), age (6–13 years or 14 years or older) and TFI (maximum TFI score for six maxillary anterior teeth). We obtained a composite measure of SES through appraisals of parents’ levels of schooling and frequency of dental attendance. We obtained the parents’ assessments of their children’s anterior esthetics from their responses to the reference question described previously that was part of a fluoride history form sent home with each subject. We focused our analysis on the 2,495 subjects for whom we had a set (parent, subject, dentist) of completed reference statements and complete data for age, sex, TFI and the reason for which the subject was dissatisfied with tooth color.

Of 2,495 subjects, 789 (31.6 percent) were dissatisfied with their tooth color, and of those subjects, 552 (70.0 percent) felt that their teeth were too yellow.

Since we incorporated TFI measurements into the model, the evaluations effectively were "controlling" for different levels in the TFI score. Starting with a fully saturated model that contained all the previously described covariates and their two-way interactions, we used a backward selection process with P < .05 as the retention criterion. We eliminated variables that did not meet this criterion from the model starting with the variables with the lowest F statistic. We describe the effects in terms of least squares means, which were the mean scores (on the 1 to 5 Likert-type scale) of the dentists, subjects and parents, adjusted for the effects of the other variables in the model. Higher scores represent disagreement with the reference statement. We contrasted the differences between pairs of least squares means using a t test. If the covariate had more than two levels, we used the Tukey-Kramer adjustment for multiple comparisons.

	RESULTS

TOP ABSTRACT METHODS AND SUBJECTS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
TFI intraexaminer reliability was a coefficient of 0.72, and interexaminer reliability was a coefficient of 0.63 (on a per subject basis). A total of 8,281 subjects were examined in the 1996–1997 school year, 49.6 percent were from the Comox/Courtenay and Campbell River communities (F-E), and 50.4 percent were from the Kamloops community (S-F). Among these subjects, 49.8 percent were girls, 50.2 percent were boys, 50.1 percent were in grades two and three, and 49.9 percent were in grades eight and nine. Children from the study sites had fairly homogeneous SES and demographic features.^34,35 Of the 2,495 subjects with parent, dentist and subject ratings and complete data for age, sex, TFI score and esthetic evaluation, 48.5 percent were boys and 51.5 percent were girls; 35.5 percent were 7 to 13 years of age and 64.5 percent were 14 to 19 years of age; 36.8 percent had a low SES and 63.2 percent had a high SES; and 62.0 percent, 19.8 percent, 13.3 percent and 4.3 percent had TFI scores of 0, 1, 2 and 3, respectively. Of the 14,970 maxillary anterior teeth examined, 1,531 (10.2 percent) had TFI scores 2, of which 728 (47.6 percent) were central incisors, 506 (33.1 percent) were lateral incisors, and 297 (19.4 percent) were canines.

Table 2 shows the subjects’ esthetic ratings for the color of their teeth (as per the reference statement) and the reasons they gave for giving the rating. Of 2,495 subjects, 789 (31.6 percent) were dissatisfied with their tooth color (disagreed or disagreed strongly with the reference statement), and of those subjects, 552 (70.0 percent) felt that their teeth were too yellow. Fluorosis was cited as a reason for dissatisfaction by 92 (11.7 percent) of the subjects who were dissatisfied with their tooth color; 71 (9.0 percent) of those who were dissatisfied were uncertain of the reason.

Of the 2,495 subjects evaluated, dentists felt that 213 (8.5 percent) subjects had unsatisfactory tooth color. They attributed this to the teeth’s being too yellow in 68 (31.9 percent) of these subjects, and they cited fluorosis as the source for dissatisfaction in 65 (30.5 percent) of these subjects.

We used repeated-measures ANOVA with a backward selection process to identify the variables associated with differences in ratings among the three categories of respondents. Covariates remaining in the model were rater (P < .0001) and two-way interactions between rater and subject’s sex (P = .0006), age (P < .0001), reason for dissatisfaction (P < .0001) and TFI score (P < .0001).

Table 3 shows the adjusted least squares means for subjects, parents and dentists (A1, B1 and C1, respectively) and t tests for subjects, parents and dentists (A2, B2 and C2, respectively) for the differences between means for each level of the covariates. Girls (A1, 3.35) were more critical (that is, had higher mean scores) than were boys (A1, 3.29), and these means were higher than those of parents of girls (B1, 2.78) and of dentists of girls (C1, 2.68). The difference between male and female subjects’ evaluations was statistically significant (A2, P = .0139), while parents (B2, P = .026) and dentists were more critical of boys’ tooth color (C2, P = .022). While younger subjects were more critical than older subjects (A2, P < .0001), parents of younger subjects were less critical than those of older subjects (B2, P < .0001). Dentists’ ratings were not significantly associated with subjects’ age group. Subjects with TFI scores of 1 and 2 were not significantly more critical than were subjects with a TFI score of zero, while those with TFI 3 were (A2, P = .009). Similarly, only parents of subjects with TFI 3 had significantly different ratings (B2, P < .0001). For dentists, there were significant differences between TFI = 2 and TFI = 3 and TFI = 0 (C2, P = .011 and P < .0001, respectively), but there was no significant difference between TFI = 1 and TFI = 0. There was no difference between the esthetic ratings of teeth with TFI scores of 1 and zero for any of the raters. Subjects, dentists and parents were most critical of enamel opacities (A1, 3.59; C1, 3.27, and B1, 3.22, respectively).

	DISCUSSION

TOP ABSTRACT METHODS AND SUBJECTS RESULTS DISCUSSION CONCLUSIONS REFERENCES

Table 2 shows each of the stakeholder’s reactions to the reference statement. First, 789 subjects (31.6 percent) disagreed with the reference statement, compared with 479 (19.2 percent) of parents and 213 (8.5 percent) of the dentists’ evaluations. So it appeared that subjects had a dissatisfaction ratio 1.7 times greater than that of parents and 3.7 times greater than that of dentists. Focusing on respondents who disagreed with the reference statement, of 789 subjects, 552 (70.0 percent) cited the tooth color being too yellow as the reason for their dissatisfaction, while 92 (11.7 percent) cited fluorosis. Of 479 parents who disagreed with the reference statement, 284 (59.3 percent) cited tooth color as being too yellow, while 60 (12.5 percent) cited fluorosis. Finally, of the 213 subjects dentists felt had unsatisfactory tooth color, dentists cited tooth color as being too yellow in 68 (31.9 percent) of these subjects, while they cited fluorosis in 65 (30.5 percent) of these subjects.

In addition to having a lower threshold for dissatisfaction than dentists, subjects’ and parents’ perceptions of tooth color’s being too yellow were their predominant reasons for dissatisfaction compared with that of dentists. Not only was fluorosis not a major issue for any of the stakeholders, but similar numbers of dentists and parents cited fluorosis as the reason for dissatisfaction.

Interestingly, some of our original speculations about each group’s expectations and standards proved to be incorrect. SES did not meet the P < .05 retention criterion, and we excluded it from the model. The results of the ANOVA were consistent with the hypothesis that girls were more critical about the color of their teeth than were boys, but parents of boys were more critical than were those of girls, contrary to our hypothesis. While we hypothesized that dentists would be more critical of the esthetic appearance of girls’ teeth than that of boys’ teeth, our data suggest the opposite. The adjusted mean score for boys was 2.74, while the adjusted mean score for girls was 2.68 (P = .0005). This finding is in conflict with that of Beyer and Lindauer,⁴ who found that parents and dentists were somewhat less tolerant of midline deviations in female subjects than in male subjects (P < .026). Even though both studies address an obvious esthetic issue, perhaps we are tapping different priorities or domains within appearance when midline location in particular and overall appearance in general are being appraised.

Our data did not support the hypotheses that older children are more critical of the color of their anterior teeth than are younger children or that dentists were more critical of older children’s anterior teeth esthetics. In fact, younger children were significantly more critical. Parents of older children were, as we hypothesized, more critical than parents of younger children.

Our model yielded several interesting contrasts across the three groups of evaluators. We start by qualifying the infrequency of identical agreement situations as the least likely outcome because there were five categories. If entirely random, we would only expect 4 percent (2 percent x 2 percent) to agree exactly. Probability caveats aside, the subjects in all strata were substantially more critical of the color of their anterior teeth than were parents or dentists. Generally speaking, lay observers tend to be more stringent in their perceptions of esthetic appearance than are dentists. There are several reasons to believe that, even among professionals, a gradient exists that connotes more exacting standards to conform to esthetic norms. Among some of the manifestations of the gradient, Levy and colleagues⁴⁰ compared dental students’ perceptions of dental fluorosis and enamel opacities in their first and fourth years of dental studies and found that "in general the [fourth-year] students were more likely to report favorable scores than they were in their first year." Cochrane and colleagues¹ found that orthodontists and maxillofacial surgeons were more likely to identify images with a Class 1 relationship as being most attractive than were dental students or laypeople. Kokich and colleagues² found that general dentists and laypeople were less able to detect discrepancies than were orthodontists. Ellwood and colleagues¹³ speculated that such differences might be a result of professional training and experience. In the specific case of esthetic assessments of fluorosis, the shift was "a form of dental esthetic conditioning that influences the objectivity of the scoring." We assume that in this context, Ellwood and colleagues meant that dentists are better able to identify small variations from the physiological norm than are laypeople. However, one issue would be the ability to identify a condition, and an entirely different one would be to consider such a condition worth mentioning as important.

Our interpretation of the phenomenon that dentists appear to be less critical of mild-to-moderate degrees of fluorosis is threefold. First, dentists are able to identify the nuances of fluorosis and other variations in tooth color,^18,20,21,23 and they can diagnose them. In fact, Stephen and colleagues²⁵ attributed an almost 11 percentage-point difference between lay observers’ and dentists’ identification of teeth blemished with diffuse mottling to "the untrained eye of the lay observer [being] inevitably prone to overestimate mottling." Second, dentists are less inclined to see fluorosis as an esthetic problem than are lay observers, who usually are the people directly involved in the situation other than professionals—for example, parents or children with the condition.^18,23 Third, the fact that other dental conditions falling in the same category of esthetic concerns (for example, tooth alignment, about which orthodontists are more critical than patients^3,4,9 or parents^3,4) are nevertheless ascribed different importance by dentists suggests that there is a loosely defined ranking of conditions. Some of these other conditions are perceived by professionals to be more serious than others, but mild fluorosis does not appear to be one of them.

Dentists may regard fluorosis with less concern than do first-year dental students because they have been exposed to a greater diversity of oral conditions in their clinical practice.

Rational explanations account for this placement of mild fluorosis toward the lower end of the hierarchy of conditions. First, dentists associate fluorosis with exposure to fluoride intake during tooth mineralization, which will increase resistance to caries. Dentists could see this sequel as a natural consequence of a desirable preventive measure. Second, as Levy and colleagues⁴⁰ have suggested, dentists may regard fluorosis with less concern than do first-year dental students because they have been exposed to a greater diversity of oral conditions in their clinical practice. Albino and colleagues⁴¹ attributed discrepancies between patients’ and dentists’ esthetic perceptions to the dentists’ understanding of the problem and of the limited potential for improvement of the situation. Dentists are cognizant of the limitations and trade-offs relating to treating the condition, while patients focus on the issue of appearance.⁴¹ In this context, the relative importance of mild fluorosis to patients could diminish until it ceases to be seen as an esthetic problem worth noting, let alone fixing.

	CONCLUSIONS

TOP ABSTRACT METHODS AND SUBJECTS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
The perception of anterior occlusion differs from that of tooth color. While the literature shows that dentists are consistently more critical of tooth shape and arrangement than are patients^4,5,9,10 or their parents,^5,9,10 they are less critical of tooth color than are patients or parents.^20,26 Moreover, children are more critical of their tooth color than parents or dentists, with girls being more critical than boys, and with the majority of children feeling that their teeth are too yellow. Parents and children see white teeth as esthetically pleasing, despite the dentist’s evaluation. Fluorosis, at the level generally prevalent in North America, is not perceived by parents or children as being a major esthetic issue.

The three stakeholders in the esthetic treatment of child and adolescent patients—parents, dentists and patients—appear to see the potential outcome of such treatment differently. This has two practical consequences. First, dentists should not assume that what they identify as an esthetic problem would be seen as such by parents or patients. Conversely, parents and patients may identify an esthetic problem that a dentist does not consider critical. For example, our data suggest that there likely is an untapped demand for bleaching in pediatric practices. Second, patients and parents may disagree with dentists and each other that the result of treatment is esthetically acceptable. This reinforces the importance of dentists’ ensuring that all three stakeholders are in agreement about the course of treatment, the rationale for undertaking it and the expected results.

View larger version (121K): [in this window] [in a new window]   Dr. Shulman is a professor and the graduate program director, Department of Public Health Sciences, Baylor College of Dentistry, Texas A&M University Health Science Center, 3302 Gaston Ave., Dallas, Texas 75246, e-mail "jshulman{at}tambcd.edu". Address reprint requests to Dr. Shulman.

View larger version (133K): [in this window] [in a new window]   Dr. Maupomé is an investigator, Kaiser Permanente Center for Health Research, Portland, Ore.

View larger version (114K): [in this window] [in a new window]   Dr. Levy is a professor and the graduate program director, Department of Preventive and Community Dentistry, College of Dentistry, University of Iowa, Iowa City, and a professor, Department of Epidemiology, College of Public Health, University of Iowa.