Back to search

University of Crete (UOC)

Εκτίμηση της ηλικίας από την μεταμόρφωση της κλείδας

Abstract

dc:description

Introduction: Estimating a person's age is of extreme importance in forensic science. When a person is prosecuted and does not carry official documents certifying his age, as is can be the case with illegal immigrants, one’s age must be determined, as one has a different sentence if one has reached the age of adulthood. Furthermore, in forensic death cases, including mass disasters and decomposing corpses, age estimation is crucial for reconstructing the biological profile and, thus, the identification of the individual. When finding unidentified skeletal material, there are four questions that must be answered: age, sex, ancestry, and stature. The human skeleton undergoes age-related changes over time. Epigrammatically, childhood is characterized by dental development, adulthood is characterized by the completion of skeletal growth, and old age is characterized by progressive deconstruction. The value of the clavicle in the age estimation is because its proximal/sternal epiphysis is the last epiphysis of the long bones to ossify. The proximal clavicle is completely ossified, usually between 25 and 30 years of age. Thanks to the long development period, the clavicle is a good bone material for age estimation in young adults. The existing studies are based on the gradual maturation of the proximal clavicular epiphysis until its complete ossification, with its evaluation macroscopically, with the help of radiology and microscopically. In Greece, the clavicle has only been studied as far as gender is concerned, and there is no reference study that could be used to estimate age in Greeks. Aim of the study: The study aims to investigate the usefulness of the clavicle in estimating the age of the adult after the closure of the epiphyses of the bone. The transformation of the sternal end of the clavicle is studied in a modern Greek population using morphological and imaging methods. The material of the study comes from the osteological Cretan collection but also from CT (computed tomography) scans of patients that were recalled as anonymous data from the archive of PAGNI (University General Hospital of Heraklion). The study's successful outcome will significantly contribute to the database of modern Greeks and the Forensic Science community by proposing a new age estimation method for the Greek population. As part of this study, a comparative study is also carried out on osteological material from Thailand and even a small sample from Athens. Material and methods: 1) The macroscopic part of the study. The sample of clavicles from the Cretan Collection has been studied macroscopically. These are clavicles of 95 individuals, 43 women, and 52 men, from the Cretan Collection, which consists of bones from the Cemetery of St. Konstantinos and Pateles in Heraklion, Crete, and was created in 2005. The population of the collection consists mainly of Cretans or people who lived in Crete for more than three generations and who lived and died between the end of the 19th century and the beginning of the 20th.The population selection criteria used are:1. good condition of the clavicles2. representation of all ages for both sexes3. absence of injuries or pathological changes in the bones4. confirmed Cretan origin.5. known age and gender from existing death certificates or ossuary inscriptions. The average age for women is 73 and for men 67. All and every clavicle of the sample was documented with many photographs taken. The changes in the sternal end of the clavicles but also the entire surface of the clavicle were described and grouped in two ways: A) Following the protocol of Falys and Prange 2015, which consists of up to 6 stages for the following three characteristics: topography, porosity, and osteophyte formation on the sternal end of the clavicle. B) Creating a second prototype and more detailed descriptive protocol. Specifically, at the sternal end of the clavicle, the following characteristics were studied and grouped:• morphology, shape, relief, and porosity of the articular surface• special features of the relief (pits and fossa)• margins and the morphology of the lip• presence of groove and nodule of ossification• epiphyseal plate and gap between the epiphyseal plate and bony surfaces• scar from the union of the ossification nuclei and • morphology of the notch for the first rib.2) Comparative study and application of the method for age estimation based on the degeneration of the sternal end of the clavicle in the specimen from Greece and a bone collection specimen from Thailand. The comparative study and application of the protocol of Falys and Prange 2015, which consists of up to 6 stages for the following three characteristics topography (Topography), porosity (Porosity) and osteophyte formation (Osteophyte formation) in the clavicle in a sample of clavicles of 72 individuals (38 men, 34 women, aged 22–84) from the bone collection in Chaing Mai, Thailand in collaboration with Pongpon Traithepchanapai. The Greek sample also included 7 male samples from Athens. Then, for each clavicle, the cumulative score for these three characteristics was obtained, and a comparison was made with the age range of the corresponding cumulative scores of the above study. In addition, the mathematical formula for age estimation from the clavicle based on the two morphological characteristics (topography and porosity, with the sex given or unknown) was applied to both samples. The error between observers and the main observer was also calculated. Kappa statistics (Cohen, 1968) and regression equations from the original study were used in our sample and subsamples separated by gender and population. The correlation of each morphological characteristic and the composite score with age was evaluated using Spearman's correlation coefficient (rs). 3) Estimation of the age from the metamorphosis of the sternal end of the clavicle with the help of computed tomography in the Greek population. In this part of the study, clavicles from 196 CT scans of patients recalled as anonymous data from the PAGNI archive. The case search was conducted through Evorad with a search term CT scan Thorax between 2014 and 2020.This became possible after an application and granting a relevant license by decision 354 of the 8/29-04-2020 meeting of the scientific council of PAGNI. Anonymized patient data were used. The patient's age had to be between 10 and 40 years, and both sexes, male and female, aimed to be equally represented in the sample population. CT scans from 87 women and 109 men were thus examined. An additional exclusion criterion was axial ST thickness less than or equal to 1 mm. As a result, the CT scans in question were 0.625mm or 1.0mm thick. Concurrent coronal reconstructions were available in most cases, but the thickness of the slices was variable, between 1.0- and 5.0 mm. Age was known in all cases as date of birth, but social/economic status or other additional information was unavailable. All selected patients had Greek surnames. In each patient, the morphology of the sternal end of the clavicle was assessed on all available CT scans via Evorad, and its developmental stage was recorded separately for the left and right clavicles, respectively. The ossification of the sternal epiphysis of the clavicle was based on the 5-stage classification of Schmeling et al. but also with the extensive, more detailed staging of stages 2 and 3 described by Kellinghaus et al. Results:1) The macroscopic part of the study. A) Following the protocol of Falys and Prange 2015. The results of applying the Falys and Prangle 2015 method to the Crete sample are considerably more imprecise than those reported in the original publication. Specifically, the application of the method resulted in 55-58% of the total sample (aged 19-101) being classified within a 95% confidence interval. The percentage increases if the method is applied only to ages >40 years to 61-64%. The classification increases when the method is used to left clavicles even though it was developed for right clavicles. Besides, the method performs only 1% more when the gender is known. These results are less valid and accurate than those reported in the original publication (blind tests composite score and regression formula in ages >40 years resulted in 96.4% for both methods of the total sample being classified within a 95% confidence interval), justify the modification of the protocol to create a new method for the data of Crete. B) Creating a second prototype and more detailed descriptive protocol. The next step was the analysis of the new data on the population of Crete. From the analysis of our data carried out in collaboration with Papadimitriou Vasilios, MSc student in Statistics & Operational Research, The University of Edinburgh, School of Mathematics, the following conclusions emerged:• The variables with the highest correlation with age (p<0.05) for the left sternal end of the clavicle are the relief of the articular surface, the margins, and the porosity of the articular surface. •Of all variables for the left acromial end, only three variables (articular surface relief, porosity, and margins) showed a relationship with age. In contrast, for the right acromial end, articular surface morphology highly correlated with age. The special features of the relief (fossae and pits) were correlated with sex.n •Regression models were created to classify the samples into over or under 65 years. Ac-cording to univariate analysis, the best models use the following variables: the margins of the articular surface and the scar from the union of the ossification nuclei.• The results of this study show that the morphological characteristics of the clavicle are highly related to age. Regression models have shown promising results in separating the sample into individuals over and under 65. However, this observation has limited forensic importance. Further analysis needs to be done with a wider sample to provide reliable age estimation equations for the Greek population.2) Comparative study and application of the method for age estimation based on the degeneration of the sternal end of the clavicle in the specimen from Greece, and a bone collection specimen from Thailand. Correlation of all characteristics (topography, porosity, and osteophyte formation) with actual age was found in both Greek men and Greek women (p<0.05). However, this was not the case in Thai men, where no statistically significant association was observed with any variables or the composite score for the left clavicle. In Thai female subjects, no statistically significant correlation was observed for the right clavicle and osteophytes for the left clavicle. The results of Spearman's rank correlation coefficients in the Greeks are in accordance with Falys and Prangles' original study, which also showed a correlation between all three traits and the composite score. Bilateral differences were also examined with paired t-tests, and no statistically significant differences were found between the right and left sides except for osteophyte formation for the overall sample (p<0.01), the Greek sample (p<0.016) and the Greek female sample (p<0.033).For each value of the composite score, a t-test was performed for age comparing males and females, with results showing a trend for the mean age of female skeletons to be greater than that observed for male skeletons for the same composite score except for the composite score 8, 10 and 11 for the right side and 7, 9 and 16 for the left side. After a Bonferroni correction for multiple testing, p-values indicated no significant difference between males and females in composite scores for both sides except for a score of 14 for the left side (p<0.003). Suppose the populations in our study are considered separately. In that case, the same trend is observed: the mean age of women tends to be older than the mean age of men for the same composite scores, especially as the score increases for both populations. There is a tendency for the Greek sample to have higher values compared to the Thai sample for both men and women. The Greek sample provided more accurate estimates for both sides than the Thai sample. The known-gender regression study gave slightly better results in most cases. When people under 40 were excluded, the classification increased for both Thais and Greeks. There is a greater overestimation of age in the <40 age groups that improve in the 40-80 age groups. In contrast, all subgroups show age underestimation (negative bias) for people over 80. However, it should be emphasized that the original study did not include people under 40. However, another related study, Price's study, included them. Interobserver error study reveals a fair to moderate interobserver agreement, which could be attributed to the limited experience of the second observer. 3) Estimation of the age from the transformation of the sternal end of the clavicle with the help of computed tomography in the Greek population. Preliminary statistical analysis of the data using R had the following results: Spearman's correlation analysis revealed a positive correlation between age and stage of clavicle ossification in both male and female subjects. Stage 2c ossification was first observed at 18 years in female subjects (N=6) and at 20 years in male subjects (N=3) when the left clavicle was examined. Correspondingly at 18 years for female subjects (N=5) and at 19 years for male subjects (N=5) when the right clavicle was examined. Stage 3c ossification was observed in adult subjects (N=25) in male and female subjects, except for one subject considered an outlier. Afterward, a stage transition analysis was carried out according to the Bayesian model. Probability density functions were calculated using informative priorities for age distribution in the total population, deceased and violently deceased individuals. Our study showed that when the medial clavicular epiphysis’ ossification is sorted in stages 4 and 5 in Greek males and females, the cumulative probability of adulthood (≥18 years) is nearly 1. As far as Greek males are concerned, in stage 3c, the cumulative probability of adulthood (≥18 years) is 0.86, and in Greek females in stage 3c, the cumulative probability of adulthood (≥18 years) is 0.97. This is the first study of age estimation based on the ossification of the sternal clavicular end using CT in a Greek population. Conclusions: • The results of the study of age estimation from the clavicle macroscopically, both in the material from the Cretan collection and in the material from Greece and Thailand, with the method of Falys and Prange 2015, indicate that the morphological characteristics of the clavicle are highly related with age.• These results, however, are less valid and accurate than those reported in the original publication (blind tests of composite score and regression formula in ages >40 years resulted in 96.4% for both methods of the total sample being classified within 95% confidence interval) and justify the modification of the protocol to create a new method for the Cretan data.• Regression models from the prototype method have shown promising results in separating the sample into people over and under 65. However, this observation has limited forensic importance. Further analysis needs to be done with a wider sample to provide reliable age estimation equations for the Greek population.• Regarding the inter- and interobserver error for macroscopic measurements, the study's results agree that the accuracy and reliability of any age estimation method are closely related to the observer's experience when applying the method. Moreover, even in the study by Blom et al and Price, reproducibility is limited regarding composite scores after intra- and inter-observer testing. Both Falys et al. and Blom et al. have noted the need for better descriptions of the method with 3D examples. In our study, there is rich photographic material for each clavicle separately.• Regarding the estimation of the age from the clavicle with the help of CT scans and the ossification classification of the sternal epiphysis of the clavicle based on the 5-stage classification of Schmeling et al but also with the extensive, more detailed staging of stage 2 and 3 described by Kellinghaus et al, the results of this study indicate that when the medial clavicular epiphysis is scored in stages 4 and 5 in both Greek males and females, the cumulative probability of adulthood (≥18 years) is nearly 1. This is in accordance with a previous systematic review by Hermetet et al., which describes that people in stages 4 and 5 are at least 18 years old. However, as far as Greek males are concerned in stage 3c, the cumulative probability of adulthood (≥18 years) is 0,857, while according to Hermetet et al., males in stage 3c are also at least 18 years old. Hermetet et al. state that caution is required at the same stage for women. In our study of Greek females in stage 3c, the cumulative probability of adulthood (≥18 years) is 0.968.

Degree

thesis:*
Grantor dc:publisher
University of Crete (UOC)
Year dc:date
2025

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Madentzoglou, Maria
  • Μαδεντζόγλου, Μαρία

Subjects

dc:subject × 20

Rights

Language dc:language
gre

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/59011

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Madentzoglou, Maria; Μαδεντζόγλου, Μαρία. Εκτίμηση της ηλικίας από την μεταμόρφωση της κλείδας. University of Crete (UOC), 2025. http://hdl.handle.net/10442/hedi/59011