Introduction

In recent years, age estimation has been applied to many aspects of forensic cases [1]. The chronologic age estimation in the range of 14−23 years, however, remains a complex problem [2]. The ages of 14−18 years are important for determining legal responsibility in Chinese law [1,2,3,4]. During these years, however, it is difficult to identify the chronologic age exactly.

The method of age estimation includes hand−wrist plain radiographic examination, cervical vertebrae plain radiographic examination, and sexual maturity [5]. These methods are limited, however, by epiphysis−diaphysis measurements and physical development, which could be affected by numerous factors, including economic status, assets, and living environment.

Dental age estimation provides significant references in forensic identification [3]. Most teeth are mature by about 14 years of age [4]. The mineralization and maturity of the third molars occur later than in other teeth. The development of the third molar is quite prolonged. Many previous studies have confirmed that third molar mineralization is available for age estimation [2, 6,7,8,9,10].

The shape, timing of formation, and eruption of the third molar vary according to the region of origin, ethnic group, and sex [11]. To estimate the age of an ethnic group, especially those 14−18 years old, it is first necessary to take into consideration the changes in mineralization of the third molars. So far, there has been no research reported regarding the Han ethnic group in Hunan Province in southern China.

Demirjian’s classification is the most widely used for evaluating children and adolescents at present [9, 12, 13]. In this study, we analyzed the mineralization of the third molars using Demirjian’s classification and made statistical comparisons about the subject’s sex, age, and location of the third molar. Two formulas were devised to estimate age based on mineralization stage and sex.

Materials and methods

Our hospital’s ethics committee approved the study, and orthopantomograms were obtained only from patients who had given written consent for the use of their data according to the Declaration of Helsinki. The data were collected for 2519 patients (1190 males, 1329 females; 8−23 years of age). The orthopantomograms were obtained at XiangYa stomatological Hospital Central South University in Hunan Province during 2012−2016. The chronological age was calculated from the birth date and the digital orthopantomograms date and converted to the age with two decimal places. Inclusion criteria were the following: (1) southern Chinese and Han who were born and living in Hunan Province, (2) no history of medical or surgical disease that could affect the eruption or mineralization of teeth, and (3) well-nourished and without any chronic disease or genetic disease. Exclusion criteria included the following: (1) orthopantomograms of inadequate quality that were difficult to be evaluated; (2) cysts or tumors in mandible and/or maxillae; (3) dental pathology affecting dental development, macrodontia, microdontia, and supernumerary; (4) those who had no third molars or who had a history of third molar extraction;(5) the family history of genetic disease that could affect the eruption or mineralization of teeth. The patients’ sex and age distribution are shown in Table 1.

Table 1 Number of the individual groups by ages and genders
Full size table

On the basis of Demirjian’s eight-stage classification [11, 12], the mineralization of third molars was assessed at nine stages. Stage 1 was added and indicated the appearance of a third molar bud but without calcification [14,15,16]. Two professional examiners evaluated the mineralization stages. In all, 500 orthopantomograms were selected randomly to test the assessment consistency between the two examiners. Two weeks after the first evaluation, the two examiners reevaluated these orthopantomograms to test rater reproducibility. The kappa test was used to analyze the results.

Formulas were devised to estimate age based on mineralization stage 1, A–G, and sex in the group who has the third molars of either side with multilinear regression analysis. Another separate sample consisted of 395 orthopantomograms (180 males, 215 females; 8–23 years of age) which had four third molars was collected for the validation of formulas using Friedman’s M test. The mean difference and absolute mean difference between predicted and chronological ages were calculated.

The data for third molar mineralization were recorded and analyzed using SPSS version 20.0 for Windows (IBM, Armonk, NY, USA). The statistical analysis was performed using the χ2 test, Student’s t test, multilinear regression analysis, and analysis of variance (ANOVA). A value of P < 0.05 was considered to indicate statistical significance.

Results

It showed a good agreement that the kappa values were 0.968 and 0.942 for inter-observer and intra-observer [17].

The mean age of the 2519 patients was 16.64 ± 4.47 years, and there was no significant difference in the number of males and females (P = 0.845, Table 1). The frequency of the stages of third molars is explicit according to Tables 2, 3, 4, and 5. According to Tables 2, 3, 4, and 5, stage E and stage G could be references to evaluate the age of a person. The person will most likely to be equal or greater than 12 years old, when a third molar reaches the stage E. The person will probably be over age 14, when a third molar reaches the stage G. The mean ages and standard deviations of the different mineralization stages in males and females are shown in Table 6. The mean ages of the initial mineralization and the crown completion were around 9.66 and 13.88 years in males and 9.52 and 14.09 in females, respectively. The minimum ages of apical closure were around 16 years in both sexes. There was a significant difference 28 at stage C occurred 0.65 years earlier in males than in corresponding females. Thirty-eight and 48 at stage F occurred 0.76 and 0.84 years earlier in males, respectively, than in corresponding females. Twenty-eight at stage G occurred 0.44 years earlier in males than in corresponding females. In other stages, there was no difference found between males and females.

Table 2 Frequency of the modified Demirjian’s stages of tooth 18 in both sexes
Full size table
Table 3 Frequency of the modified Demirjian’s stages of tooth 28 in both sexes
Full size table
Table 4 Frequency of the modified Demirjian’s stages of tooth 38 in both sexes
Full size table
Table 5 Frequency of the modified Demirjian’s stages of tooth 48 in both sexes
Full size table
Table 6 Statistic data of chronological age in third molars
Full size table

The comparison of four third molars in Demirjian’s eight stages in males and females is shown in Table 7. There was a significant difference between mandibular and maxillary teeth at stage C in males. Mineralization of 18 occurred 0.8 years earlier than 38 and 1.07 years earlier than 48 at stage C. Mineralization of 28 occurred 0.95 years earlier than 38 and 1.22 years earlier than 48. There were no significant differences for all eight stages in females.

Table 7 Statistic data of Demirjian’s stages of different genders
Full size table

According to the third molar and sex, a multilinear regression analysis was used to analyze the individuals who had two third molars on either right or left side to establish the age model. Multilinear regression parameters were calculated and shown in Table 8. The model based on the mineralization stages and sexes is shown in Table 9.

Table 8 Multiple regression results for calculating the age with gender and stages of third molars
Full size table
Table 9 The age prediction model was based on the mineralization stages and sexes
Full size table

Additionally, the total of new 395 patients who had four third molars on both sides had selected to check the reliability of the regression equation. The inclusion criteria and exclusion criteria are referred to “Materials and methods.” The mean difference using tooth 18 and 48 was − 0.02 years and using tooth 28 and 38 was 0.02 years. Absolute mean difference values were 0.13 and 0.13 years, respectively. There was no significant difference found between actual age, predicted age of 18 and 48, and predicted age of 28 and 38 (Table 10).

Table 10 Reliability of age estimation by means of the derived formula
Full size table

Discussion

The process of mineralization and maturity of the third molars occurs later than those of the other permanent teeth. When estimating the age of unidentified juveniles, the development and mineralization of the third molars are very important, especially when the other permanent teeth have completed the development [2, 6,7,8, 13]. Many methods have been used to analyze tooth development and eruption, such as those described by Demirjian et al. [12, 13], Thorson and Hägg [9], and Olze et al. [18]. Demirjian’s classification has been widely used in the world, and its stages are easily applied to, and modified for, most ethnic groups [12, 13].

The mineralization stages for ages 14, 16, and 18 were the most meaningful for application to actual cases [1,2,3,4]. The age of the patients in the present study was 8−23 years. The minimum age limit was set according to previous studies [15, 19,20,21]. A high level of consistency between the mineralization stage and age has been proved [13, 16, 19, 22].

Previous studies have studied the relation between the development of third molars and chronologic age. Zandi et al. [23] studied 2536 orthopantomograms of Iranian individuals between 5 and 26 years of age. They found that the first appearance of a third molar bud was about 9 years at average, and the mean ages of complete crown formation were around 14 years. Jung et al. [21] analyzed 2490 patients of Korean between 6 and 24 years of age. They found that the age of initial mineralization of the third molars was around 8.75 years, and the average age of crown completion was 14.52 and 15.04 years of the maxillary and mandibular, respectively. Streckbein et al. [24] evaluated 2360 orthopantomograms of German and found that all individuals at stage H of all third molars were over 18 years. For central southern Chinese Han in Hunan Province, the crown completion was around 14 years and the initial age of apical closure was around 16 years. There was no meaning to calculate or compare the mean/maximum age of stage H, because the stage of third molars could not move on if it reached the stage H [25]. The differences showed that the third molars vary among ethnic groups, regions, and generations and between the sexes [11]. And the diversity could also be influenced by dietary habits and lifestyles.

China has a large territory and 56 different ethnic groups, which could be the reason of the differences in the mineralization of third molars in different areas of China. Zeng et al. [8] studied 3100 patients aged 4.1–26.9 years in Guangdong Province, China. The results showed the ages of complete crown formation were 13.6 years in males and 13.8 years in females. The minimum ages of the apical closure were 18.1 years in males and 17.6 years in females. In their study, the mineralization of 28 occurred 0.25 years earlier than 38 in females at Demirjian’s stage C. In males, at Demirjian’s stage G, they found the mineralization of 38 and 48 was earlier than 28 and 18, respectively, which was different from the conclusion that maxillary mineralization was earlier than that of mandibular teeth in several stages in our study and other studies [20,21,22,23]. At stages A to H, mineralization was earlier in males than that in females, whereas in our study, it showed no difference in most stages. Guo et al. [14] studied 2078 patients aged 5–23 years in Sichuan Province, the western of China. The age of third molars reached complete crown calcification was around 14 years, which is similar with our study and other research results [14, 20, 25]. Bai et al. [22] evaluated 291 orthopantomograms from the patients aged 10–25.2 years in central China. The results showed the average ages of complete crown formation of third molars were 13.1 years in males and 14 years in females. The minimum ages of apical closure were 22 years in males and 19.6 years in females. Qing et al. [20] analyzed 2192 orthopantomograms from southwestern China. They found that there was no significant difference in mineralization between males and females except for molar 48 at stage E, and the mineralization of third molars in the maxillary region was earlier than that in the mandibular region. Comparing our study with these studies, the age of complete crown formation is similar with that of the other studies. But the initial age of root apical closure is earlier than the males and females of central Chinese and southern Chinese in Guangdong Province. Most of the studies concluded that there was no significant difference between both sides, and males were earlier than females in the age of crown completion and apical closure, which was consistent with our studies [8, 14, 16, 20,21,22,23, 25,26,27]. But some studies from other countries found there was no significant difference between males and females [28,29,30]. More studies and a great number of patients may be needed to reach a meaningful statistical conclusion.

The formula presented will help the practicing forensic odontologist to estimate the chronologic age around 8–20 years based on the mineralization of third molars. Considering there was no significant difference between both sides in maxillary and mandible and the effect of multicollinearity, we chose the third molar on the left or right side to establish the multiple linear regression formula. It has been proven that each independent variable in the formula has a significant effect on age estimation. Although it has been tested and validated and could be used in age estimation, it will be more precisely in combination with other teeth or body structures to estimate the chronologic age, especially in the individuals who do not meet the evaluation criteria [16, 27, 30]. Hofmann et al. studied the third molars and educed formulas of either the maxilla or the mandible and proved there was no significant difference between the median predicted age and the median actual age [31].

Conclusion

This study represented a comprehensive analysis of third molars of a Han population in central southern China and provided references for the age estimation of juveniles and adolescents in this area by the mineralization of third molars according to the Demirjian’s classification. The equation could be used to estimate the chronologic age of individuals who have two third molars on either the right or the left side.