Background: Mesiodistal and buccolingual measurements of the teeth have been the traditional means of determining gender which sometimes are difficult to measure.
Aims and Objectives: To assess the degree of sexual dimorphism in upper permanent canines and first molars using diagonal diameters and to evaluate the applicability of diagonal measurements in sex determination.
Materials and Methods: Mesiobuccal-distolingual (MBDL) and distobuccal-mesiolingual (DBML) measurements of the crown and cervix of maxillary canine and maxillary molar were taken using digital vernier callipers in 100 upper dental casts obtained from 50 males and 50 females of the age group of 17-25 y.
Results: With step wise discriminant function analysis, results were statistically significant with males showing higher diagonal measurement values. The classification accuracy for males was 69% and in females was 73% with overall accuracy of 71%. Results showed that canines exhibited highest accuracy with cervical measurements being more dimorphic. MBDL measurements were found to be more reliable. Most dimorphicmeasurements included right molar cervical MBDL, left molar cervical MBDL followed by right and left molar crown MBDL.
Conclusion: In cases of tooth rotations, crowding, certain fillings and attritions where traditional mesio-distal and bucco-lingual measurements are not applicable, diagonal measurements could be of help in predicting gender.
Introduction
Gender determination has been the focus of many forensic studies and is of significance in mass fatality cases where bodies are damaged beyond recognition. Further, in incidents where only jawbone fragments with teeth are found, gender determination is possible only with the help of teeth. Teeth being resistant to postmortem destruction and being chemically stable are an excellent material for odontologic and forensic investigations [1]. Therefore, teeth can be used as a tool of reconstructive identification particularly in cases of major catastrophes [2].
Determining gender using dental features is mainly achieved by comparing the tooth dimensions in males and females using traditional mesiodistal and buccolingual crown diameters of teeth. But these dimensions get affected by attrition, interproximal wear facets, cervical abrasions, crowding, and presence of dental calculus in cervical third. Further, these dimensions are difficult to measure when the teeth are still held in socket [2,3]. As a result, alternative measurements including the crown and cervical diagonal iameters were developed. According to Hillson et al., most ofthe archaeological and fossil remains of human teeth exhibit heavy attrition of usual crown diameters but alternative diagonal measurements are not affected until later stages of wear. Exclusion of contact points remains one of the major advantages of diagonal diameters of molars. The diagonal diameters proposed by Hilson et al., included Mesiobuccal-distolingual crown diameter, Mesiolingualdistobuccal crown diameter, Mesiobuccal-distolingual cervical diameter and Mesiolingual-distobuccal cervical diameters [4].
Canines being least affected by periodontal disease are last to be extracted. Canines have been recovered from dental remains even in hurricanes and air disasters and are found to withstand extreme conditions [5]. Boaz et al., has mentioned sexual dimorphism of canine with larger crowns in males, because of longer period of amelogenesis [6].
Maxillary first molars are among the early permanent teeth to erupt at an age of 6-7 y. Accuracy of gender determination in mature individuals is possible if the postcranial skeleton is intact. But in case of younger children, gender estimation from skeleton is difficult. In those cases, odontometric features of teeth that erupted at an early age can be of immense use in gender determination according to a study on the Haryana population by Sonika V [1].
Therefore, aim of the present study is to determine the accuracy with which gender can be differentiated using odontometric diagonal measurements in Davangere population. The objective of the study is to assess the degree of sexual dimorphism in upper permanent canines and first molars using diagonal diameters and to evaluate applicability of diagonal measurements in sex determination.
Materials and Methods
Study sample included 100 upper dental casts [Table/Fig-1] obtained by alginate impression after informed consent, from 50 males and 50 females of the age group of 17-25 y in a population of Davangere, in Karnataka. Diagonal measurements of crown and cervix of upper canines and first molars were taken.
Using the study dental casts, mesiobuccal-distolingual and distobuccal-mesiolingual measurements were taken separately. All the measurements were taken from both right and left sides using a 6” digital vernier calliper of 0.01mm resolution.
When placing the calliper parallel to the occlusal surface [Table/Fig-2], the following points were taken as guide during the measurements as defined by Hillson et al., [4].
MBDL : the largest distance between the mesiobuccal corner and the distolingual corner of the crown
DBML: the largest distance between the distobuccal corner and the mesiolingual corner of the crown
Mesiobuccal–distolingual cervical diameter (cervical MBDL ) is defined as the maximum distance from the mesiobuccal corner of cemento–enamel junction point to the distolingual corner.
Mesiolingual–distobuccal cervical diameter (cervical MLD B) is defined as the maximum distance from the mesiolingual corner of cemento–enamel junction point to the distobuccal corner.
Two crown and two cervical diagonal measurements of upper canines and first molars were taken. Each dimension was measured twice at a different time interval by the same observer. Selection criteria included fully erupted permanent maxillary canines and maxillary first molars of 17-25 y age group subjects. Extractions, fillings, crowns or orthodontic apparatuses, orthodontic anomalies and teeth with dental caries involving the point angles (in canines) and cusp tips (in molars) that could affect the odontometric measurements were excluded from the study.
Results were subjected to discriminant function analysis (using SPSS software statistics). Step wise discriminant function analysis allows a researcher to test whether a set of independent variables can be used to discriminate group membership by generating a valid equation. Wilks’ lambda is the criterion for tooth variable selection. It is used to add or remove variables from the analysis according to their discriminatory power. It denotes how useful a given variable is in the step wise analysis and determines the order in which variables enter the analysis.
Results
Four diagonal measurements on each tooth (including two cervical and two crown), altogether 16 measurements on both sides, were obtained in each of the individuals. Each measurement was measured twice at two different intervals by the same examiner to check for intra observer error. There was no statistically significant difference between the first and second measurements made as shown in [Table/Fig-3].
All the 16 measurements taken revealed greater values in males than in females and the differences were statistically significant. Highly significant dimorphic measurements included right molar cervical MBDL followed by right molar crown MBDL, left molar crown MBDL and left molar cervical MBDL. All the values on right side of jaw were greater than the left side except for molar crown and cervical DBML values [Table/Fig-4,5]. Overall classification accuracy of results was 71% with males exhibiting 69% and females exhibiting 73%, with accuracy rate being higher in females [Table/Fig-6]. Accuracy rate was higher in canines, as shown in [Table/Fig-7] and MBDL measurements were found to be more reliable. Cervical measurements revealed to be more dimorphic as seen in [Table/Fig-8].
Stepwise discriminant function analysis
Discriminant function analysis [Table/Fig-9,10] was carried out to find which of the variables affected the functions the most. This analysis allows the examiner to test whether a set of independent variables can be used to discriminate group membership and provides the equation for such discrimination.
Discriminant function score (Z) for gender assessment = -17.665-0.32(RtCanCrMBDL) + 0.509(RtCanCrDBML) + 0.796(RtCanCerMBDL)+0.038(RtCanCerDBML)+0.109(RtMolCrMBDL)-0.904(RtMolCrDBML)+ 0.695(RtMolCerMBDL)+0.664(RtMolCerDBML)-0.399(LtCanCrMBDL)+0.669(LtCanCrDBML)-0.217(LtCanCerMBDL)-0.869(LtCanCerDBML)+0.411(LtMolCrMBDL)-0.162(LtMolCrDBML)-0.180(LtMolCerMBDL)-0.094(LtMolCerDBML).
If the discriminant score is below sectioning point, individual is classified as female and if the score is above sectioning point, individual is classified as male.
Discussion
Teeth being the most durable tissue in the body, have been the focus of interest for various forensic odontogenic studies [3].
A variety of factors contribute to the magnitude of dimorphism in the teeth. Environmental influence due to variation in food resources of different populations can be the major cause. Further, there can also be complex interaction between genetic and environmental factors resulting in variation in the magnitude of dimorphism in the teeth. Garn et al., has proposed that teeth through the course of evolution behave in many ways ranging from reduction of the entire dentition to reduction of one group of teeth in relation to other [7].
According to Kalia S, difference in the balance of hormonal production between the sexes consequent to the differentiation of either male or female gonads during the sixth or seventh week of embryogenesis attributed to the difference in sizes of teeth rather than any direct effect of sex chromosomes themselves [8].
In forensics, mesio distal and bucco lingual width measurements have been routinely used for gender determination. These measurements provide accurate results and should be the preferred method if width measurements are possible. But conditions like tooth rotations, crowding, orthodontic anomalies, occlusal attrition, Mesio-Occlusal, Disto-Occlusal, Mesio Occluso distal fillings may prevent one from taking accurate width measurements. In those situations, diagonal measurements would be of help in determining gender and also these diagonal axes do not include the contact points of the crown, and are therefore not affected by approximal attrition until the facets become large enough to include them [4]. But caution is needed in taking diagonal measurements as measurements will go wrong if the callipers is not placed correctly [9].
According to the results of the present study, the diagonal measurements were significantly greater in males than females. The results were in accordance with various other studies revealing clear dimorphic differences between male and female teeth [6,10].
According to a study by Hashim HA, Murshid ZA evaluating 720 teeth in a Saudi population, canines were the only teeth to exhibit dimorphism [11]. In a study carried out by Karaman in Turkish population, diagonal measurements of canines revealed greatest accuracy rate with MBDL measurements being more reliable, which is consistent with the present study [9].
Accuracy rate of gender determination was found to be higher in females than males according to Karaman which is also in accordance with the current study [9]. According to Bishara et al., first molars exhibited significant sexual dimorphism with respect to buccolingual dimensions [12]. Also, according to Karaman, if the accuracy rate of width measurements of certain teeth is high, then the accuracy rate of diagonal measurements of such teeth is also gh [9].
According to Hilson et al., alternative diagonal measurements were as reliable as normal width measurements and remain a promising method for sex determination. Diagonal measurements would be better measures for worn out teeth in archeological and fossil dental remains [4].
Cervical diagonal diameters of molars were found to be more sexually dimorphic than crown diagonal diameters in a study performed over modern greek population similar to that of the present study. Cervical measurements are not affected until most of the tooth is lost and thereby offer greater advantage in archeological specimen identification [4]. Analysis of the data generated reveals that diagonal measurements can be an useful aid in gender determination especially in situations where width measurements are not applicable [3]. But taking diagonal measurements has got certain limitations as the measurements are most difficult to obtain and require more attention [9]. Measurements might go wrong if the callipers is not positioned properly parallel to tooth axis.
The method applied in the present study is simple, inexpensive and can therefore be applied in forensic odontology for establishing gender identity of an individual.
In future, alternative diagonal measurements should find wider usage along with conventional width measurements and in near future, digital imaging techniques might be available to measure different aspects of tooth size and shape which are not strongly affected by wear.
Figure of ten cast samples taken together
Recording diagonal measurements of molar using digital vernier callipers
Paired t-test evaluating intra-observer variation in crown and cervical diagonal measurements of canines and molars (of both right and left sides). (None of the t-values are significant at the p<0.05 level)
| Teeth | First observation | Second Observation | ‘t’ value | ‘P’ value |
| Mean | Standard deviation | Mean | Standard deviation |
| RtCanCrMBDL | 6.95 | 0.48 | 7.02 | 0.47 | 1.83 | 0.07 |
| RtCanCrDBML | 7.18 | 0.47 | 7.22 | 0.50 | 1.53 | 0.13 |
| RtCanCerMBDL | 7.56 | 0.51 | 7.53 | 0.50 | 0.46 | 0.65 |
| RtCanCerDBML | 7.36 | 0.53 | 7.31 | 0.56 | 0.73 | 0.47 |
| RtMolCrMBDL | 12.18 | 0.63 | 12.17 | 0.62 | 0.28 | 0.78 |
| RtMolCrDBML | 10.51 | 0.64 | 10.50 | 0.66 | 0.40 | 0.69 |
| RtMolCerMBDL | 12.33 | 0.64 | 12.32 | 0.62 | 0.41 | 0.68 |
| RtMolCerDBML | 10.64 | 0.65 | 10.65 | 0.64 | 0.45 | 0.66 |
| LtCanCrMBDL | 6.90 | 0.53 | 6.91 | 0.49 | 0.43 | 0.67 |
| LtCanCrDBML | 7.14 | 0.48 | 7.13 | 0.48 | 0.41 | 0.68 |
| LtCanCerMBDL | 7.51 | 0.50 | 7.49 | 0.46 | 1.22 | 0.23 |
| LtCanCerDBML | 7.28 | 0.52 | 7.29 | 0.50 | 0.18 | 0.86 |
| LtMolCrMBDL | 12.12 | 0.62 | 12.14 | 0.60 | 1.09 | 0.28 |
| LtMolCrDBML | 10.63 | 0.58 | 10.64 | 0.60 | 0.24 | 0.81 |
| LtMolCerMBDL | 12.25 | 0.64 | 12.26 | 0.63 | 0.84 | 0.40 |
| LtMolCerDBML | 10.79 | 0.59 | 10.77 | 0.62 | 1.14 | 0.26 |
Percentage of sexual dimorphism of right side canines and molars
Percentage of sexual dimorphism of left side canines and molars
Accuracy of classification results in determining gender
| Accuracy of classification results |
| Variable | Predicted group membership | Total average (%) |
| Male (%) | Female (%) |
| RtCanCrMBDL | 59 | 53 | 56 |
| RtCanCrDBML | 63 | 58 | 60 |
| RtCanCerMBDL | 61 | 62 | 62 |
| RtCanCerDBML | 58 | 58 | 58 |
| RtMolCrMBDL | 63 | 75 | 69 |
| RtMolCrDBML | 65 | 75 | 70 |
| RtMolCerMBDL | 63 | 71 | 67 |
| RtMolCerDBML | 58 | 69 | 64 |
| LtCanCrMBDL | 49 | 52 | 51 |
| LtCanCrDBML | 60 | 56 | 58 |
| LtCanCerMBDL | 54 | 60 | 57 |
| LtCanCerDBML | 63 | 60 | 61 |
| LtMolCrMBDL | 63 | 69 | 66 |
| LtMolCrDBML | 60 | 66 | 63 |
| LtMolCerMBDL | 60 | 66 | 63 |
| LtMolCerDBML | 60 | 69 | 65 |
| Overall | 69 | 73 | 71 |
Accuracy of determination of gender using upper canines and molars
| Accuracy of classification results |
| Function | Predicted group membership | Total average (%) |
| Male (%) | Female (%) |
| Canines | 67 | 66 | 67 |
| Molars | 65 | 66 | 65 |
Accuracy of determination of gender using crown and cervical measurements
| Accuracy of classification results |
| Function | Predicted group membership | Total average (%) |
| Male (%) | Female (%) |
| Crown | 72 | 64 | 68 |
| Cervical | 69 | 68 | 69 |
Stepwise discriminant function analysis of mesiobuccal-distolingual and mesio-lingual-distobuccal crown and cervical dimensions of upper canines and molars. F value for a variable indicates its statistical significance in the discrimination between groups. Wilk’s lambda is the criterion for tooth variable selection
| Stepwise discriminant analysis of means of diagonal measurements of canines and molars for gender identification |
| Variable entered | Wilks' Lambda | Exact F statistic | df1 | df2 | Sig. |
| RtCanCrMBDL | 0.987 | 1.153 | 1 | 90 | 0.286 |
| RtCanCrDBML | 0.967 | 3.055 | 1 | 90 | 0.084 |
| RtCanCerMBDL | 0.935 | 6.292 | 1 | 90 | 0.014 |
| RtCanCerDBML | 0.962 | 3.564 | 1 | 90 | 0.062 |
| RtMolCrMBDL | 0.896 | 10.444 | 1 | 90 | 0.002 |
| RtMolCrDBML | 0.979 | 1.956 | 1 | 90 | 0.165 |
| RtMolCerMBDL | 0.877 | 12.618 | 1 | 90 | 0.001 |
| RtMolCerDBML | 0.967 | 3.035 | 1 | 90 | 0.085 |
| LtCanCrMBDL | 0.989 | 0.97 | 1 | 90 | 0.327 |
| LtCanCrDBML | 0.969 | 2.896 | 1 | 90 | 0.092 |
| LtCanCerMBDL | 0.983 | 1.513 | 1 | 90 | 0.222 |
| LtCanCerDBML | 0.982 | 1.687 | 1 | 90 | 0.197 |
| LtMolCrMBDL | 0.905 | 9.425 | 1 | 90 | 0.003 |
| LtMolCrDBML | 0.968 | 3.011 | 1 | 90 | 0.086 |
| LtMolCerMBDL | 0.913 | 8.617 | 1 | 90 | 0.004 |
| LtMolCerDBML | 0.955 | 4.232 | 1 | 90 | 0.043 |
Canonical discriminant function coefficients for diagonal diameters groups
| Variable | Unstandardised coefficienta | Structured matricsb | Standardised coefficientsc | Group centroids | Sectioning Point |
| Male | Female |
| RtCanCrMBDL | -0.688 | 0.2 | -0.329 | 0.525 | -0.503 | 0.00 |
| RtCanCrDBML | 1.091 | 0.721 | 0.509 |
| RtCanCerMBDL | 1.598 | 0.656 | 0.796 |
| RtCanCerDBML | 0.073 | 0.623 | 0.038 |
| RtMolCrMBDL | 0.181 | 0.596 | 0.109 |
| RtMolCrDBML | -1.425 | 0.509 | -0.904 |
| RtMolCerMBDL | 1.155 | 0.418 | 0.695 |
| RtMolCerDBML | 1.032 | 0.383 | 0.664 |
| LtCanCrMBDL | -0.748 | 0.355 | -0.399 |
| LtCanCrDBML | 1.421 | 0.354 | 0.669 |
| LtCanCerMBDL | -0.437 | 0.352 | -0.217 |
| LtCanCerDBML | -1.682 | 0.345 | -0.869 |
| LtMolCrMBDL | 0.696 | 0.284 | 0.411 |
| LtMolCrDBML | -0.28 | 0.264 | -0.162 |
| LtMolCerMBDL | -0.296 | 0.25 | -0.18 |
| LtMolCerDBML | -0.16 | 0.218 | -0.094 |
| Constant | -17.665 | | |
a. Unstandardised discriminant function evaluated as group means;
b. Pooled within groups correlations between disciminating variables and standardised canonical discriminant functions
c. Discriminant score less than sectioning point is catogorised as female & more than sectioning point is considered as male
Conclusion
The evolving field of forensic odontology in India depends a lot on simple and economical methods to identify persons with their dental remains and fragmented jaws. In those conditions, a dentist may be called upon to render expertise in forensic science.
In general, Mesiodistal and Buccolingual measurements are more reliable than diagonal measurements. Diagonal values are preferred over the width measurements in attrition, malpositions such as tooth rotation, crowding, and orthodontic anomalies which cause difficulties in taking width measurements. So along with width measurements, diagonal measurements of teeth can be used in determining the gender.
To conclude, crown and cervical diagonal diameters of canines and molars can be used as an aid for sex determination in support of width measurements. Thereby, they can aid in identifying a person from fragmented jaws and dental archaeological remains. This study needs further evaluation in other populations as sexual dimorphism is a population specific phenomenon.
a. Unstandardised discriminant function evaluated as group means;b. Pooled within groups correlations between disciminating variables and standardised canonical discriminant functionsc. Discriminant score less than sectioning point is catogorised as female & more than sectioning point is considered as male
[1]. V Sonika, K Harshaminder, GS Madhushankari, J Kennath, Sexual dimorphism in the permanent maxillary first molar: a study of the Haryana population J Forensic Odontostomatol 2011 29(1):37-43. [Google Scholar]
[2]. M Malik, Sex determination by diagonal teeth measurements Dissertation submitted to the Rajiv Gandhi Institute Of Health Sciences, Karnataka Bangalore 2008 [Google Scholar]
[3]. E Zorba, K Moraitis, C Elipoulos, C Spiliopoulou, Sex determination in modern Greeks using diagonal measurements of molar teeth Forensic Science International 2012 217:19-26. [Google Scholar]
[4]. S Hillson, C FitzGerald, Flinn Helen, Alternative Dental Measurements: Proposals and Relationships with Other Measurements American Journal Of Physical Anthropology 2005 126:413-26. [Google Scholar]
[5]. S Kaushal, VVG Patnaik, G Agnihotri, Mandibular canines in sex determination J Anat Soc India 2003 52:119-24. [Google Scholar]
[6]. K Boaz, G Gupta, Dimorphism in maxillary and mandibular canines in establishment of gender J of forensic Dental Sciences 2009 1(1):42-44. [Google Scholar]
[7]. SM Garn, AB Lewis, DR Swindler, RS Kerewsky, Genetic control of sexual dimorphism in tooth size J Dent Res 1967 46:963-72. [Google Scholar]
[8]. S Kalia, Study of permanent maxillary and mandibular canines and inter-canine arch widths among males and females Dissertation submitted to the Rajiv Gandhi University of Helth Sciences, Karnataka, Bangalore 2006 [Google Scholar]
[9]. F Karaman, Use of diagonal teeth measurements in predicting gender in a Turkish population Journal of Forensic Sciences 2006 51(3):630-85. [Google Scholar]
[10]. JL Anderson, GW Thompson, Interrelationship and sex differences of dental and skeletal measurements J Dent Res 1973 52:431 [Google Scholar]
[11]. MA Hashim, ZA Murshid, Mesiodistal tooth width: A comparison between Saudi males and females, Part 1 Egypt Dent J 1993 39:343-46. [Google Scholar]
[12]. SE Bishara, D Ortho, JR Jakobsen, EM Abdullah, AF Garcia, Comparisons of mesiodistal and buccolingual crown dimensions of the permanent teeth in three populations from Egypt, Mexico and the United states Am J Orthod Dentofacial Orthop 1989 96(5):416-22. [Google Scholar]