INTRODUCTION

Angle's classification of malocclusion is based on the mesiodistal relationship between the upper and lower dental arches.1 Subsequent studies revealed that the dental arch relationship is largely affected by the facial skeleton upon which the teeth are invested.234 Björk23 pointed out that local changes confined to the teeth can exercise a limited effect on the discrepancy in the relationship between the upper and lower dental arches, whereas facial or cranial configurations may produce marked effects. Thus, the relationships between craniofacial morphology and malocclusion have long been of interest to orthodontists. Studies have primarily investigated the relationship between the type of malocclusion and cranial base angle,5–7 vertical facial dimensions,8 and size of both jaws.910 However, few studies have investigated the relationship between the anteroposterior dental arch and jaw-base relationships.11–15

An area of extensive debate within orthodontics is whether the anteroposterior dental arch correlates with the jaw-base relationship, and thus whether both need to be assessed in orthodontic diagnosis and treatment planning.16 Because assessments of both the anteroposterior dental arch and jaw-base relationships are made in the sagittal plane, it is plausible that they are highly correlated, though some have suggested that there is poor relationship between them.16 A population-based study is required to strengthen the evidence as to whether the anteroposterior dental arch relationship is correlated with the jaw-base relationship.

Haavikko and Hele12 studied 1017 Finnish children aged from 2.5 to 19 years and concluded “among Finnish children every Angle class includes a large proportion of atypical facial types”. However, the age range of subjects within that study varied considerably and thus the sample had varying degrees of dental arch and jaw-base development. They also reported that in Angle Class III the frequency of skeletal Class III (38%) was smaller than that of skeletal Class I (62%).12 In contrast, Milacic and Markovic14 reported that 85% of Angle Class III individuals had identical dental and skeletal relationships in a study of 585 Swiss orthodontic patients. The conflicting evidence is likely to be related to the samples that have been studied, which for the most part have been convenient clinical samples as opposed to random samples drawn from a population. From a large sample of 8- to 16-year-old Caucasian orthodontic patients, a moderate association was reported between the first permanent molar relationship and four different methods of assessing jaw-base relationship13 ie, r = .50 for ANB and r = .64 for Wits appraisal. However, the proportion of Class II patients was more than 70%, far above that of a population sample.16

It is generally accepted that Angle's classification is valuable when assessing the anteroposterior dental arch relationship. Both angular and linear measurements have been proposed in the assessment of anteroposterior jaw-base relationship.1317–19 However, each of the methods described exhibits its own inherent weakness, based on the variability of factors other than the jaw relationship itself.1317 The correlation between anteroposterior dental arch relationship and anteroposterior jaw-base relationship assessed by angular and linear measurements has not been undertaken in a population study.

This study aims to investigate the association between anteroposterior dental arch and jaw-base relationships in a population sample of Chinese children. In addition, the study aims to compare angular vs linear measurements of the association between anteroposterior dental arch and jaw-base relationships.

MATERIALS AND METHODS

Data Collection

Study cast measurement

Two trained and calibrated examiners assessed the molar relationship of the 306 orthodontic casts (anteroposterior dental arch relationship) on the basis of Angle's definition using dividers and standard rulers. The upper first molar mesiobuccal cusp width of each subject was measured and was regarded as one cusp width. Molar Class I was defined as occurring where the mesiobuccal cusp of the upper first molar occluded with the mesiobuccal groove of the lower first molar within the range of less than half a cusp width anteriorly or posteriorly. Molar Class II was defined as occurring where the mesiobuccal cusp of the upper first molar occluded anterior to the Class I position. Molar Class III was defined as occurring where the mesiobuccal cusp of the upper first molar occluded posterior to the Class I position. Twenty-five percent of the casts were reassessed for inter- and intraexaminer reliability. The study sample of 306 subjects was distributed into three different classes based on the molar relationship (Figure 1).

View Figure

• Download as Powerpoint
• Download Figure

Cephalometric analysis

The 405 lateral cephalometric radiographs were digitized twice and analyzed by an independent examiner using the program CASSOS (Computer-Assisted Simulation System for Orthognathic Surgery; Soft Enable Technology Limited, Hong Kong, PRC). The cephalometric reference values of anteroposterior jaw-base relationship were assessed using angular measurement by the ANB angle18 and linear measurement by the Wits analysis19 of the 405 lateral cephalograms (Figure 2), and the jaw-base relationships were categorized into three groups (skeletal Class I, II, III). Skeletal Class I was defined as values within the range of mean value of the sample ± one standard deviation. Class II was defined as greater than mean value plus one standard deviation. Class III was defined as less than mean value minus one standard deviation (Table 1). Thereafter, the study sample of 306 subjects was allocated to the three different skeletal classes based on the ANB angle and Wits appraisal, respectively (Figure 1).

View Figure

• Download as Powerpoint
• Download Figure

Table 1.  Classification of Anteroposterior Jaw-Base Relationships Assessed by ANB and Wits Analysis in 405 Children

View Fullscreen

Data analysis

In assessing inter- and intraexaminer reliability, method errors (M.E.) were calculated by Dahlberg's formula,22 M.E. = Σ d²/2n, where d is the registrations differences between the two registrations, and n is the number of double registrations. The M.E. for the intra- and interexaminer's measurements of molar relationships were not statistically significant (P > .05). The M.E. for linear and angular measurements of the cephalometric radiographs did not exceed 0.5 mm and 0.7°, respectively (P > .05).

Frequency tables were produced to determine the distribution of the children with different anteroposterior dental arch relationships (Angle's definition of molar relationship) and different jaw-base relationships (ANB angle and Wits analysis). The percentage of concordance between the dental arch and jaw-base relationships was assessed. Variations in ANB angle (°) and Wits appraisal (mm) with respect to molar classification were assessed using Kruskal-Wallis tests (a nonparametric equivalent to analysis of variance). Association between molar classes and skeletal classes was assessed through cross-tabulation using Fisher's exact test. Correlation analysis was used to investigate the association between ANB angle and Wits appraisal.

RESULTS

The distribution of the study sample of 306 children into the molar and skeletal classes is given in Figure 2. The dento-skeletal morphology assessed from the lateral cephalograms did not differ between those included in this study and those with missing orthodontic study casts. Class I molar relationship was most common (73.2%), followed by Class II (19.9%) and Class III (6.9%). The jaw-base relationship (skeletal classes) assessed by the ANB angle and Wits analysis differed significantly among the three molar classes (Figure 3).

View Figure

• Download as Powerpoint
• Download Figure

Significant association between molar classes and skeletal classes was found when the jaw-base relationship was assessed by the ANB angle (P < .01), and a stronger significant association was found when jaw-base relationship was assessed by the Wits analysis (P < .001) (Table 2). Sixty-one percent of the subjects had identical anterior-posterior classes in dental arch and jaw-base relationships determined by ANB angle, 67% of subjects had identical classes determined by Wits analysis, and 53% with both the ANB angle and Wits analysis (Table 3).

Table 2.  Distribution of 306 Children With Identical Anteroposterior Dental Arch and Jaw-Base Relationships

View Fullscreen

Table 3.  Distributions of 306 Children With Various Dental Arch (Molar Classes) and Jaw-Base Relationships (Skeletal Classes) Assessed by ANB (°) and Wits Analysis (Mm) on the Basis of Molar Classes

View Fullscreen

Among the molar Class I group, 72.8% were categorized as skeletal Class I according to ANB and 78.1% as skeletal Class I by Wits analysis (Table 3). According to ANB, 29.5% of the molar Class II group was categorized as skeletal Class II, whereas according to Wits analysis 37.7% were categorized as skeletal Class II. Of the molar Class III cases, 33.3% and 38.1% were categorized as skeletal Class III by ANB and Wits analysis, respectively. In the molar Class II group, only one subject (1.6%) was categorized as skeletal Class III by Wits analysis, whereas 13.1% were categorized as Class III by ANB. Among the Class III molar group, no subject was categorized as skeletal Class II by either the ANB or Wits analyses. Two-thirds of the sample was categorized as having identical jaw-base relationships as assessed by both ANB and Wits analysis. According to both ANB and Wits analysis, 55.3% were categorized as skeletal Class I, 6.9% as skeletal Class II, and 6.5% as skeletal Class III.

The distribution of the molar relationship is given for each of the skeletal classes assessed by ANB and Wits analysis, respectively, in Table 4. Both ANB and Wits skeletal Class I had a similar proportion of molar Class I cases (76.9% vs 77.8%) and also similar distributions of molar Class II and Class III cases. Skeletal Class III with ANB had a significantly larger proportion of molar Class II (16.0%) than with Wits (2.5%). There was no skeletal Class II with molar Class III case, but the proportion of skeletal Class II with molar Class II was significantly larger (56.1%) with Wits than with ANB (40.9%).

Table 4.  Distributions of 306 Children With Various Dental Arch (Molar Classes) and Jaw-Base Relationships (Skeletal Classes) Assessed by ANB (°) and Wits Analysis (mm) on the Basis of Skeletal Classes

View Fullscreen

The correlation for the reference sample of 405 children between the ANB angle and the Wits appraisal showed that the r-value was .66 (P < .001). The correlation coefficient for the study sample of 306 children was .65, whereas for the groups of Class I, II, and III dental arch relationships it was .60, .64, and .67, respectively (Figure 4).

View Figure

• Download as Powerpoint
• Download Figure

DISCUSSION

The interrelationships between the most widely used method to assess dental arch relationship, Angle's classification,1 and the two most commonly used cephalometric methods, ANB angle18 and Wits appraisal,19 were investigated in a comprehensive population sample for the first time in the orthodontic literature. The present sample comprised study models and lateral cephalograms obtained from a large proportion of a randomized sample of 12-year-olds included in an oral and health survey in the mid-1980s.20 Almost 20 years later, all cephalograms were available, but only three out of four orthodontic study models. Some orthodontic study models were missing, and others were partly damaged and thus not amenable to analysis.

The dento-skeletal morphology assessed from the lateral cephalograms did not differ between those included in this study and those with missing orthodontic study casts, which might indicate that the “dropout” or “loss” was at random. The present sample's prevalence of molar classes was in general agreement with a large population study of Chinese children.23 The mean values of ANB and Wits are also in general agreement with previous studies.2425 Consequently, the results of this study could be considered as representative of its original population.

The choice of a 12-year-old cohort for the oral health survey was mainly due to the following reasons20: (1) the severity and the type of malocclusion that can be detected from a child at this age is representative of his/her problem in the future; (2) a better assessment of the original crown diameter of the permanent teeth can be made because little or no enamel attrition has altered the mesiodistal or buccolingual dimensions of the teeth; and (3) a child at this age would tend to have fewer carious, restored, or extracted teeth due to the recent emergence of the permanent teeth, and so would be ready for orthodontic treatment. The cephalometric values obtained from 12-year-olds seem to be valid for a larger age range because it has been reported that there is no statistically significant difference for cephalometric parameters in 7- to 13-year-olds.26

This study showed that only half of the subjects had an agreement with dental arch relationship and jaw base relationship assessed by angular and linear measurements (Table 3). However, there was an agreement between dental arch and jaw-base relationship in two out of three subjects when assessed with one method only. Consequently, the “facial type” did not match the type of occlusion in about 30% of the subjects. It is thus evident that the Angle classification of occlusion, based on the dental arch relationship alone, will not reveal the full picture of dento-facial deformity, not even in the sagittal plane. This was in accordance with other studies.1214 It is obvious that orthodontic patient samples can be heavily biased, eg, it was reported from a study using orthodontic patient samples that dental arch and skeletal Class III individuals coincide in 85% of patients, whereas results from this study and another population study27 demonstrate that the unbiased figure is far less (Table 2).

This study showed that Wits appraisal based on linear measurement along the occlusal plane seems to reflect the dental arch relationship and the underlying jaw-base relationship better than angular measurement (ANB angle) using the same landmarks on the maxilla (A-point) and mandible (B-point), respectively. Consequently, the method errors for locating those two cephalometric landmarks were identical for both ANB angle and Wits appraisal. To assess the ANB angle, the nasion (N-point) was located. It has been pointed out the clinical experience has revealed that the cephalometric landmark nasion (N) is difficult to identify on Southern Chinese patients. As point N is not involved in Wits appraisal, all errors associated with its identification, as well as the inconsistencies in the angle ANB due to varying horizontal and vertical positions of N, are irrelevant.1128 Another plausible reason as to why Wits analysis proved superior might be that when defining the functional occlusal plane in Wits analysis, two landmarks are used, the midpoints of overlap of the mesiobuccal cusps of the first molars and the buccal cusps of the first premolars.29 The anteroposterior dental arch relationship using Angle's classification is based on the first permanent molars, which were assessed close to the functional occlusal plane.

The regression between the ANB angle and the Wits appraisal was statistically significant. The association is evident, since the analyses measure the same structures in a very large series. However, the correlation coefficient (r-value), which represents the probability to predict the variables from one another, was relatively low (.65). This means that a certain value of the ANB angle may be associated with many values of the Wits appraisal, and thus the prediction between variables is moderate, and the association between the ANB angle and Wits appraisal did not differ among the various classes of dental arch relationship (Figure 4C–E)

CONCLUSIONS

• The anteroposterior dental arch and jaw-base relationships did not match in one of three individuals.

• Linear measurement of anteroposterior jaw-base relationship is a more valid reflection of the dental arch relationship than the angular measurement.

• The prediction between the variables of ANB angle and Wits appraisal is moderately independent of the dental arch relationship.