Evaluation of mixed dentition analyses in north Indian population: A comparative study.

INTRODUCTION: Mixed dentition regression equations analyses (Moyers, Tanaka-Johnston) are based on European population, reliability of these methods is questionable over other population.
MATERIALS AND METHODS: The present study was conducted on total 260 study models. This study was done in two phases. In the first phase, linear regression equations were made. In the second phase, comparison of actual values of sum of mesiodistal width of canine, first and second premolars with the predicted values proposed by Moyers, Tanaka-Johnston, and the new proposed mixed dentition analysis for North Indian population were made.
RESULTS: Set of four linear regression equations for predicting sum of mesiodistal width of permanent canine, first premolar and second premolar in North Indian population from sum of mesiodistal width of mandibular incisors and mandibular first molars, were proposed as; (a) for males, maxillary arch, Y = 2.9 + 0.40X, (b) mandibular arch Y = 3.91 + 0.37X (c) for females, maxillary arch Y = 0.56 + 0.45X (d) mandibular arch Y = 1.14 + 0.42X. Moyers and Tanaka-Johnston, mixed dentition analysis, is found to be overestimating the mesiodistal width of unerupted canine and premolars in North Indian population.

Introduction

Prediction of mesiodistal tooth dimension of the unerupted permanent canines and premolars are the essential part of diagnosis and treatment planning for accurate estimation that allows the dentist to better manage tooth/arch length discrepancies in mixed dentition period.[1] There are three basic approaches that have been used to estimate the combined mesiodistal width of unerupted canine and premolars in mixed dentition stage.[2]

(1) Measurement of the unerupted teeth on radiographs. (2) Use of regression equations. (3) Combination of regression equation and radiographs, in which first and third approach of mixed dentition analyses uses the radiographs as suggested by many authors[34567] carry considerable accuracy but it is time consuming, requires an expensive setup and a trained operator which makes these prediction method difficult to be applied to assertain the exact mesiodistal widths of unerupted canine and premolars. Moreover its technique sensitive, distortions of the image and misinterpretation of rotated tooth in their crypts.[8] So these are not reliable accurate methods. Whereas, use of regression equations, mixed dentition analysis can be done with minimum systemic error, ease of use by any person, less time consuming and with minimum armamentarium required.[910]

Purpose of this study was to evaluate a new proposed mixed dentition analysis and to identify the possible sexual dimorphism in North-Indian population by use of regression equations and also to compare it with the most widely used mixed dentition analyses.[1011]

Materials and Methods

The present study was conducted on total 260 study models (i.e. 130 subjects). The impression were taken those subjects randomly come to the OPD of the Institute. This study was done in two phases, First phase of the study consisted of 160 study models (i.e. 80 subjects) of North–Indian subjects to evaluate the mixed dentition analysis and to formulate the new proposed linear regression equations for maxillary and mandibular arches in male and female separately by correlating the mesiodistal width of canines and premolars with various combinations of mandibular incisors and mandibular and maxillary first molars in the age group of 14-22 years which were collected randomly from the pre-treatment patient records and from the patients attending the OPD of the Institute.

The criteria for the selection of samples were decided according to the previous studies done by Tanaka and Johnston,[11] which were as follows:

The subjects were of North Indian background for at least one prior generation, that is both the parents had to be of North Indian origin

The subjects of sample were in the age group of 14-22 years with a mean age of 18 years, because at this age permanent canine and premolars usually erupt completely and to preclude any discrepancies based on significant proximal wear

Complete eruption of permanent incisors, canines, premolars and first molars in maxillary and mandibular arches

The teeth measured were free of proximal restorations, fractures, and proximal caries

There was no evidence of hypoplasia or anomalies of the teeth

No developmentally missing or supernumerary teeth

Subjects with severe crowding were excluded to facilitate the accurate measurement.

Digital caliper with ground tip (Baker RD-10), of accuracy 0.01 mm were used for accurate measurement and to avoid eye fatigue. The mesio-distal width of a tooth was obtained by measuring the maximum distance between contact points on the proximal surfaces. Digital caliper was held parallel to the occlusal surface if the tooth appeared to be in normal alignment as described by Hunter and Priest.[12] Otherwise, the mesio-distal crown diameter was obtained by measuring between the points where contact with the adjacent tooth would normally occur.[8] Values obtained for the right and left posterior segments were averaged so that there would be one value for the maxillary canine and premolars and one value for the mandibular canine and premolars.

The study models of first phase (160 study models), were divided into Group I males (n = 80) and Group II females study models (n = 80), which were further subdivided into maxillary (Sub Group Ia and IIa, n = 40) and mandibular study models (sub Group Ib and IIb, n = 40) [Table 1].

Table 1

Distribution of 160 study models of first phase

The linear regression equations (Y = A + B [X]) were made, where Y is a dependent variable and X is an independent variable while A and B are constants, which can be used clinically for the evaluation of combined mesiodistal width of unerupted canine and premolars.

The study models of second phase (100 study models), were divided into Group III males (n = 50) and Group IV females study models (n = 50), which were further subdivided into maxillary (sub Group IIIa and IVa, n = 25) and mandibular study models (sub Group IIIb and IVb, n = 25) [Table 2].

Table 2

Distribution of 100 study models of second phase

Comparison of actual values of sum of mesiodistal width of canine, first premolar and second premolar in both maxillary and mandibular arches as taken from study models with the predicted values proposed by Moyers,[910] Tanaka and Johnston,[11] and the new proposed mixed dentition analysis on the North Indian population were made.

Moyers[910] used the sum of the widths of the four mandibular incisors to predict the combined width of canines, first premolar and second premolar of maxillary and mandibular arches for male and female separately at 50% probability level by use of probability chart. Probability level, of 50% was taken for more precise prediction as recommended by Moyers.

Tanaka and Johnston[11] predicted the combined mesiodistal width of unerupted canine and premolars with the help of two linear regression equations of maxillary and mandibular arches. He found that the size in millimeters of unerupted canine and premolars can be predicted by taking half the width of mandibular incisors and adding 11.0 mm for maxillary teeth and 10.5 mm for the mandibular teeth.

Observations

Data were summarized as mean ± standard deviation and ± standard error with range (minimum and maximum) in Tables 3 and 4. Correlation among all studied variables was done by Karl–Pearson correlation coefficient. Relative association of dependent variable with independent variables was done by simple linear regression analysis. A two tailed (α =2) probability P < 0.05 was considered to be statistically significant, P < 0.01 as highly significant and P > 0.05 the not significant. MS-Excel (MS Office 97-2007) and Statistica (version 6, Chicago, US) were used for the analysis.

Table 3

Descriptive statistics (mean±SD) of dependent variables

Table 4

Descriptive statistics (mean±SD) of independent variables

In the first phase of the present study new proposed equations were made by correlating two dependent variables (1) sum of mesiodistal width of mandibular canine and premolars (2) sum of mesiodistal width of maxillary canine and premolars with four independent variables, (1) sum of mesio-distal width of mandibular incisors. (2) Sum of mesio-distal width of mandibular incisors and mandibular first molars. (3) Sum of mesiodistal width of right and left mandibular first molars and (4) sum of mesiodistal width of right and left maxillary first molars in Table 5 for male and in Table 6 for female subjects.

Table 5

Simple linear regression equations for each independent variable of males (Group I)

Table 6

Simple linear regression equations for each independent variable of females (Group II)

In the second phase of study the new proposed equations along with two mixed dentition analyses by Moyers and Tanaka-Johnston was used for the prediction of width of permanent canine and premolars which were compared with actual measurements of combined mesiodistal width of permanent canine and premolars from the study models in Table 7 for male and in Table 8 for female subjects.

Table 7

Comparison of three mixed dentition analyses by RSS for males (Group III)

Table 8

Comparison of three mixed dentition analyses by RSS for females (Group IV)

Discussion

Commonly used regression equations were based on European population (Moyers and Tanaka-Johnston). A considerable amount of literature is available which says that the tooth size varies from population to population and the same prediction equations cannot be used for different populations as shown in the study of Al-Khadra,[13] 1993, Schirmer and Wiltshire[14] in 1997. Similar studies were done in over twelve different populations[15161718192021222324] and the common finding was that there is considerable difference in tooth size in various populations with sexual dimorphism.

So far there has been no data reported for North Indian population, although it is densely populated. Hence the purpose of this study was to evaluate new proposed prediction method and to identify the possible sexual dimorphism in North Indian population by use of regression equations and also compare it with most widely used mixed dentition analyses (Moyers and Tanaka-Johnston).

Among four set of regression equations by each independent variable, the equation by second independent variable “sum of mesiodistal width of mandibular incisors and mandibular first molars” showed strongest association with highest coefficient of determination, minimum standard error and maximum regression ANOVA F-ratio, (P < 0.01) Figure 1a and b and Figure 2a and b. Suggesting it to be most reliable for the predictions among independent variables for the dependent variable (sum of mesiodistal width of canine, first premolar and second premolar).

Figure 1

New proposed regression equation line (bold line) of sum of mesiodistal width of maxillary (a) and mandibular (b) canine, first premolar and second premolar from sum of mesiodistal width of mandibular incisors and mandibular first molars in males with 95% confidence for β (broken lines). Scatter points denotes predicated value for each observed value

Figure 2

New proposed regression equation line (bold line) of sum of mesiodistal width of maxillary (a) and mandibular (b) canine, first premolar and second premolar from sum of mesiodistal width of mandibular incisors and mandibular first molars in females with 95% confidence for β (broken lines). Scatter points denotes predicated value for each observed value

New proposed regression equation line (bold line) of sum of mesiodistal width of maxillary (a) and mandibular (b) canine, first premolar and second premolar from sum of mesiodistal width of mandibular incisors and mandibular first molars in females with 95% confidence for β (broken lines). Scatter points denotes predicated value for each observed value [Figure 2].

New proposed regression equation line (bold line) of sum of mesiodistal width of maxillary (a) and mandibular (b) canine, first premolar and second premolar from sum of mesiodistal width of mandibular incisors and mandibular first molars in males with 95% confidence for β (broken lines). Scatter points denotes predicated value for each observed value [Figure 1].

Although second set of four regression equations (by independent variable “Sum of mesiodistal width of mandibular incisors and mandibular first molars”) is selected as new proposed regression equations for mixed dentition analysis, but regression equations made by other independent variables is also statically significant and hence can be use for prediction of combined width of unerupted canine and premolars where ever clinical situation doesn’t permit the use of second independent variable that is, Sum of mesiodistal width of mandibular incisors and mandibular first molars.

In the present study after formulating the new proposed regression equations of males and females for both maxillary and mandibular arches separately, for North Indian population, the second phase of study was done for evaluating and comparing it with the most widely used mixed dentition analyses by Moyers[910] and Tanaka and Johnston.[11]

In the second phase of study the new proposed equations along with two widely used mixed dentition analyses by Moyers and Tanaka-Johnston was used for the prediction of width of permanent canine and premolars which were compared with actual measurements of combined mesiodistal width of permanent canine and premolars from the study models.

Because data is distributed on both side of the regression equation line (positive and negative) for better comparison of three mixed dentition analyses residual sum of square (RSS) test was done. Better the line fits, smaller will be the RSS value [Figures 3–6].

Linear graph of maxillary archin males from Moyers, Tanaka and Johnston and new proposed also showed new proposed the best because its all predicted value (n = 25) scatter nearest to the actual [Figure 3].

Conclusions

Based on the results of the study following conclusion were drawn from the study:

Set of four linear regression equations for predicting sum of mesiodistal width of permanent canine, first premolar and second premolar in North Indian population from Sum of mesiodistal width of mandibular incisors and mandibular first molars, are as follows; (a) For males, in maxillary arch, Y = 2.9 + 0.40X, (b) For males, in mandibular arch Y = 3.91 + 0.37X (c) For females, in maxillary arch Y = 0.56 + 0.45X (d) For females, in mandibular arch Y = 1.14 + 0.42X

New proposed mixed dentition analysis is better for prediction of sum of mesiodistal width of canine, first premolar and second premolar than the commonly used mixed dentition analyses (Moyers and Tanaka-Johnston) in North-Indian population, when compared with actual measurements from study models

Moyers[910] mixed dentition analysis according to 50% probability levels of prediction chart, is found to be overestimating the mesiodistal width of unerupted canine and premolars in North Indian population

Tanaka-Johnston[11] mixed dentition analysis, overestimates the mesiodistal width of unerupted canine and premolars in North Indian population by 3.2 mm per arch in males and 4.0 mm/arch in females.