Forensic Science International 236 (2014) 196.e1–196.e6

Forensic Anthropology Population Data

Estimating age at death using the sternal end of the fourth ribsfrom Mexican males

Jessica Ines Cerezo-Roman a,b,*, Patricia Olga Hernandez Espinoza c

a Pima County Medical Examiner Forensic Science Center, 2825 E. District Street, Tucson, AZ 85714, USAb School of Anthropology, The University of Arizona, 1009 E. South Campus Drive, Tucson, AZ 85721-0030, USAc Centro INAH Sonora, Jesus Garcıa final s/n, Col. La Matanza, Hermosillo, Sonora 83000, Mexico

A R T I C L E I N F O

Article history:

Received 30 April 2013

Received in revised form 21 September 2013

Accepted 30 December 2013

Available online 10 January 2014

Keywords:

Forensic science

Osteology

Age estimation

Sternal end of the ribs

Mexican population

A B S T R A C T

The indicators proposed by Is can et al. (1984) are said to reflect age changes that occur in the sternal end

of the fourth rib. These indicators have been used to estimate age-at-death in adult skeletal samples.

However, Iscan et al. developed their methods using a forensic sample from Florida (U.S.A.). In order to

test the reproducibility of those methods we evaluate its accuracy for the fourth ribs by applying it to a

sample of known age and sex but of different biological affinity: modern males from Mexico City. We

found that the method developed by Is can et al. underestimates age-at-death in the Mexican sample.

Published by Elsevier Ireland Ltd.

Contents lists available at ScienceDirect

Forensic Science International

jou r nal h o mep age: w ww.els evier . co m/lo c ate / fo r sc i in t

1. Introduction

Estimation of biological age-at-death is one of the moreimportant parts of the analysis of modern and ancient humanskeletal remains. For adults, age-at-death estimates frequentlyutilize multiple indicators that reflect standard processes of bonedeposition, remodeling, and reabsorption that occur throughoutthe life of an individual. However, these processes are affected andinfluenced by numerous genetic, environmental, and culturalfactors. Taking these into consideration, the selection of appropri-ate methods for estimating age needs to be informed by method-specific data on the margins of error for each target sample.

Skeletal maturation processes provide a basis for estimating theage of a skeleton. In younger subadults, the estimation of ageusually relies on bone and tooth maturation. However, there aresubstantial variations in the timing of these developmentalchanges among different individuals, even those who do notsuffer from any major growth disruptions and/or stress episodes[1,2]. With mature adults, estimations of age-at-death are mainlyderived from evaluating degenerative processes usually caused bynormal wear and tear on the body over time. Researchers have

* Corresponding author at: Pima County Medical Examiner Forensic Science

Center, 2825 E. District Street, Tucson, AZ 85714, USA. Tel.: +1 520 248 5856.

E-mail addresses: cerezorj@email.arizona.edu (J.I. Cerezo-Roman),

patriciaolga.hernndezespinoza@gmail.com (P.O. Hernandez Espinoza).

0379-0738/$ – see front matter . Published by Elsevier Ireland Ltd.

http://dx.doi.org/10.1016/j.forsciint.2013.12.044

observed and analyzed these changes in specific skeletal samples,and they have developed classification methods to estimate group-specific age-at-death (for a summary and discussion of thesemethods and techniques see references [2–6]). These degenerativeprocesses also reflect an individual’s life history of growth,development and genetic predisposition. Considering this, it islikely that degenerative changes will differ in timing and manneramong different populations. Taking into consideration anindividual’s life history of growth, development, lifestyle, andgenetic predisposition, researchers argued for critically evaluatingexisting methods that estimate the age-at-death among differentpopulations [7–10]. Several studies have concluded that it isessential to assess the accuracy of and, if necessary, modify existingmethods to more effectively estimate the age-at-death ofindividuals from different populations around the globe [e.g., 7–10]. These types of studies are very useful as they facilitate theprocess of adjusting existing methods to specific populations andof acquiring a deeper and wider understanding of human variation.

Most standards used to estimate age-at-death were developedwith samples from North America, such as the Terry collection atthe Smithsonian Institution in Washington, D.C., the Hamann–Todd collection in Cleveland, Ohio, and individuals from theKorean War, among others [11]. These collections are primarilycomposed of Americans with Northern European and Africanancestry. Individuals in these collections had very different lifestyles and genetic heritages than Latin American populations. Thisresearch evaluates the applicability of methods developed to

J.I. Cerezo-Roman, P.O. Hernandez Espinoza / Forensic Science International 236 (2014) 196.e1–196.e6196.e2

estimate age-at-death in adults using the sternal end of ribs byIs can et al. [12,13] using a sample from Florida and Is can and Loth[14] with a modern Mexican population from Mexico City. Weevaluate the accuracy of the method, including consideration of theindividual components of the rib.

The chest plate, including the sternal end of the rib, is an areawhere age changes have been documented through time by manyresearchers using radiography and multislice computed tomogra-phy [e.g., 15–19], histological analysis [e.g., 20–24], and osteologi-cal studies [e.g., 12–14, 25, 26]. For example, McCornick andStewart [18], using roentgenograms to analyze age changes in theentire chest plate, suggested the alterations that occurred in thisarea included progressive ossification in the costal cartilages at thesternal rib end, both centrichondrally and peristernally, and oftenwith age and sex distinctive patterns. In addition, they mentionedthat other changes included maturation of the newly formed bonewith trabeculation formation, loss of the smooth contour of thecosto-manubrial junction, cupping of rib ends, osteoporoticchanges, and arthritic changes in the sternal head of the clavicles.At a gross anatomy level, the sternal end of the rib is probably oneof the areas of the chest plate that has received most of theattention in physical anthropological studies to estimate age basedon techniques proposed by Is can et al. [12,13] and Is can and Loth[14].

The technique proposed by Is can et al. [12,13] and Is can andLoth [14], which estimates the age-at-death based on degenerativeprocesses, evaluates changes that occur with age in the sternal endof the fourth rib. The authors proposed nine phases of degenerativeprogression graded from zero to eight [12,13]. They tested theprecision of this technique and found that morphological changesoccurring in the osseous rib segment have significant statisticalrelationships with the age-at-death. In these tests, there wasminimal interobserver error and estimated phase associationsdeparted from the expected by no more than one phase [14]. Is canet al. [12,13] also mentioned that factors such as physical activityand heavy labor, endocrine disorders, chronic lung disease, druguse, sex differences, diet, and intercostal variations [e.g., 16, 27–28]can affect the normal aging patterns characteristic of the rib ends.They also cautioned that the intercostal variation, positiveidentification of the fourth rib and differences between sex werethe most important factors to consider when using this method.

Other investigations, such as those from Dudar [29] and Saunderset al. [8], tested the validity of the techniques of Is can et al. [12,13]and Is can and Loth [14] and corroborated the usefulness of thistechnique to estimate age-at-death. However, Saunders et al. [8]found that the ages were usually underestimated and as ageincreased, the biases and inaccuracies also increased. Recent studiesby Fanton et al. [10] suggested that analyses using the sternal end ofthe rib showed poor reproducibility and repeatability due todifficulties in measurement, imprecision in describing pit depth, andfailure to take into account the pit shape, rim, and wall configuration.Hartnett [30] also tested the accuracy of age-at-death estimationusing the sternal end of the fourth rib. She suggested that there aresignificant differences in the observed versus actual ages andsignificant interobserver variation. She also concluded in her studythat the most important departures from the original study by Is canet al. [12,13] are the incorporation of bone quality and density asthey play bigger roles in phase assignment and revisions.

Application of these age indicators also identified differencesbetween sexes and populations [e.g., 14, 31–33]. Other researchershave modified the technique for use with ‘‘white’’ females [14,31–33]. Is can et al. [13] and Loth [34] later applied their methods to‘‘white’’ and ‘‘black’’ population samples from the United States,and found statistically significant sample differences. Loth [34]attributed those findings to differences in osseous density betweenthe two groups with each having different biological histories. Loth

[34] also noted that processes of deterioration occurred later in lifeon the sternal end of the rib in a 16th–18th century cemeterypopulation from Spitalfields, UK. Subsequently, Is can [35] didanother study on ‘‘white’’ and ‘‘black’’ males and females toevaluate discrepancies with the technique. He found that themorphological characteristics defining the phases were age-related. Also, he found significant differences in the rate andpattern of the metamorphosis by both sex and race.

In 1998, Yavuz et al. [36] studied a sample of males and femalesfrom Istanbul in order to evaluate the Is can et al. [12,33] method.They found that the phases proposed by Is can et al. [12] presentsimilar morphological characteristics to their sample. Theyconcluded that this technique can be accurately applied to Turkishpopulations. In 2000, Oettle and Steyn [37] replicated the studydone by Is can et al. [12] using a sample of ‘‘South African blacks’’from the Gauteng Province. This study concluded that theestimated ages were acceptable, but that the indicators werenot precise. They subsequently adjusted the criteria by modifyingthe existing phases and proposing new phases. Meena et al. [38]also tested the accuracy and bilateral variation of the methoddeveloped by Is can et al. [12] on a sample of Indian males andfemale from Lady Hardinge Medical College, New Delhi, India. Theyfound that the average of an individual’s phase score was notsignificantly different in either the right or left rib and concludedthat the method can be used in Indian populations with littlevariation.

Most aging standards only estimate age-at-death up to 50 yearsof age. However, using the sternal end of ribs, age can be estimatedin individuals who are older than 50 [12]. The method also hasbeen applied to historic, prehistoric, and hominid specimens[12,34,39]. Unfortunately, due to its fragile nature, this portion ofthe rib is infrequently preserved in archeological samples. It alsohas been suggested that the ribs most suitable for estimating age-at-death are the first four and the last rib [12,35,40,41]. However,Yoder and Ubelaker [42] tested this technique on the left and rightsecond through ninth ribs. They found that the fourth throughninth left ribs did not vary significantly from the other ribsanalyzed. Only the second right rib was found to vary significantlyfrom the other four ribs. They recommended the use of a compositescore to assign ages in a more accurate manner. Nikita [43] alsoexamined intercostal and age differences in the sternal rib endmorphology of documented female skeletons from Spitalfields andSt. Bride Church, London. The morphology of this area wascaptured using three-dimensional morphometric analysis, andstatistical manipulation was employed. She found statisticallysignificant differences between the fourth and the other ribs, withexception of the third rib. In addition, she found that all thecharacteristics of the ribs varied with age. However, Nikita [43] didnot find any statistically significant differences among the variousage groups and was unable to use any discriminant or multilinearregression analyses based on digitized coordinates, bringing intoquestion the rigorousness of the method for estimating the ageusing the sternal end of the ribs.

2. Materials and methods

In the current study, the analysis sample consists of 71 maleswho were more than 16 years old at the time of death. Allindividuals in the sample lived in Mexico City and were Mexican.The sample comprises a total of 55 individuals from the ServicioMedico Forense (SEMEFO) and 16 individuals from the civilcemetery San Nicolas Tolentino, located in the delegacıon

Iztapalapa (Table 1). However, not all individuals available fromthese two sample sets were selected for this study. Selectioncriteria focused on Mexican males of known age-at-death who didnot present bone-altering pathological conditions.

Table 1Known age ranges.

Age ranges Male

21–30 years 14

31–40 years 12

41–50 years 15

51–60 years 7

61+ 23

Total 71

Table 3Known age descriptive statistics.

Real chronological age

N Valid 71

Missing 0

Mean 49.58

Std. error of mean 2.232

Median 45.00

Mode 25a

Std. deviation 18.807

Variance 353.705

Skewness .384

Std. error of skewness .285

Kurtosis �.813

Std. error of kurtosis .563

Range 77

Minimum 21

Maximum 98

Sum 3520

a Multiple modes exist. The smallest value is shown

J.I. Cerezo-Roman, P.O. Hernandez Espinoza / Forensic Science International 236 (2014) 196.e1–196.e6 196.e3

Changes in the sternal end of the fourth ribs were evaluatedbased on morphologic and metric characteristics of the costochon-dral cavity, particularly the depth, shape, and wall and rimconfigurations without prior knowledge of the known age of theindividual. Measurements and observations used in this researchfollowed the procedures of Is can et al. [12] who divided these datainto three components with each component comprising a series ofstages (Table 2). The pit depth was measured with a depth calipercalibrated to 0.1 mm. The caliper was held perpendicular to thebase of the pit and the measurement was taken where the distancebetween the base of the pit and the adjacent anterior or posteriorwall was the greatest. The method stages are a progression fromzero to five where stage zero represents ribs that have justcompleted maturation processes and are fully developed. Stageone represents the beginning of degenerative changes andsubsequent stages represent increasing degenerative changes(e.g., young adults should be closer to stages 0 and 1 while olderadults should be at stages 4 or 5) (see Table 2).

3. Results

Statistical analyses were performed using the softwareprograms SPSS 16.0 and Microsoft Excel 2007. Analytical

Table 2Is can et al. [12] components and stages.

Component 1: Pit depth

0. Flat to slightly billowy extremity with no indentation (pit) �1.1 mm.

1. Definite pit formation with a depth ranging from 1.1 to 2.5 mm.

2. Pit depth ranging from 2.6 to 4.5 mm.

3. Pit depth ranging from 4.6 to 7.0 mm.

4. Pit depth ranging from 7.1 to 10.0 mm.

5. Pit depth �10.1 mm.

Component 2: Pit shape

0. Juveniles and adolescents with no pit formation at the

flat or billowy articular surface.

1. A shallow, amorphous indentation (pit) is present.

2. Formation of a V-shaped pit with thick walls.

3. Pit assumes a narrow U-shape with fairly thick walls.

4. Wide U-shaped pit with thin walls.

5. Pit is still a wide U-shape, yet deeper, more brittle, and

poorer in texture with some disintegration of bone.

Component 3: Rim and wall configuration

0. Specimens with a smooth regular rim and no wall formation.

1. Walls becoming apparent with a thick, smooth regular rim.

2. Definitely visible thick and smooth walls with a scalloped or slightly

wavy rim.

3. A transitional stage between the regularity of Stage 2 and irregularity

of Stage 4. Scalloped edges are disappearing and walls are thinning, yet

walls remain fairly sturdy without significant deterioration of

bone texture.

4. Rim becomes sharper and increasingly irregular with more frequent

bony projections, often most pronounced at the cranial and caudal

margins of the rib. Walls show further thinning and are less

sturdy with noticeable deterioration in texture.

5. The texture shows extreme friability and porosity. Rim is very

sharp, brittle, and highly irregular with long bony projections.

Occasionally, as the depth of the pit increases, openings form

in areas where walls are incomplete.

evaluations included descriptive statistics, one-way analysis ofvariance (one-way ANOVA), and analysis of bias and inaccuracies.

The sample presents a known age mean of 49.58 and a standarddeviation (SD) of 18.807. This sample presents a slight kurtosistoward older ages (Table 3). In order to understand better how therelationships between the components and the different stages inthe Mexican sample comparisons were made between the meanknown ages and the stages of each component (Table 4) and thesewere compared with similar analyses using the fourth rib from theyounger sample evaluated by Is can et al. [12]. Table 4 lists thedescriptive statistics of the males in our Mexican sample and fromIs can et al. [12]. In our Mexican sample, changes associated withthe process of aging began to manifest in ribs three and four arounda mean age of 44 years old (Table 4). This result contrasts withthose of Is can et al. [12], in which the components and initial stagespresent at mean age of 20.3 years (Table 4). When the mean datafor the fourth ribs of the Mexican sample are compared to the meanage of Is can et al. [12], the Mexican sample means are older in allphases of the components.

The second statistical analysis was a one-way analysis ofvariance (One-factor ANOVA) (Table 5). This analysis describesvariance in the pit depth (component one), pit shape (componenttwo), rim and wall configurations (component three), anddependence of variance in the known age based on comparisonof the mean values. ANOVA did not reveal a significant relationshipbetween known age and the fourth rib pit depth (Table 5). The pitshape reveals a level of 0.06, and, therefore there is no statisticalsignificance between the known age and the pit shape (Table 6).However, rim and wall configurations (component 3) have asignificant relationship (P = 0.01) with the known age (Table 7).

The analysis of bias and inaccuracy (Table 8) were madefollowing current studies such as Hens et al. [9], Saunders et al. [8],and Santos [44]. The bias is the mean over- or under-prediction,S(estimate age � known age)/n, where n = the number of cases.The inaccuracy is the average absolute error of age estimation,without reference to over- or under-prediction, Sjestimateage � known agej/n. The means were estimated as the mid-pointof age category ranges except for the last open-end category where61 was used. The results of the analysis of bias and inaccuracyreveal that the degree of bias and inaccuracy generally is higher asage increments. There is a lower bias in age estimation up to age 30.There is a shift from slight underestimation of age to a higherdegree of underestimation after age 40. Age estimations over age60 are dramatically underestimated. Fig. 1 graphically displays theresults of known age and estimated age. Each individual is plottedon the graph so that the degree of bias and inaccuracy may bevisualized. Bias and inaccuracy in this study were then compared

Table 4Fourth rib descriptive statistics.

Components, stages and fourth rib SEMEFO & Tolentino Is can et al. (1)

N Mean age SD N mean age SD

Component 1: Pit depth

0. Less than 1.1 mm – – – – – –

1. 1.1–2.5 mm 6 44 24.44 9 20.3 3.32

2. 2.6–4.5 mm 32 47 17.44 29 30.7 12.40

3. 4.6–7.0 mm 16 55 17.71 31 40.9 13.72

4. 7.1–10.0 mm – – – 9 55.0 15.39

5. 10.1 mm or more – – – 4 57.5 12.92

Total 54 49 18.51 82 37.9 16.15

Component 2: Pit shape

0. No pit formation, flat or billowy surface – – – – – –

1. A shallow, amorphous indentation (pit) 4 38 14.52 4 17.3 0.50

2. V-shaped pit with thick walls 12 40 19.62 15 22.8 3.28

3. Narrow U-shape with fairly thick walls 19 52 17.25 28 30.5 9.61

4. Wide U-shaped pit with thin walls 14 51 16.35 22 47.1 11.61

5. Wide U-shape, and poorer in texture 5 64 19.04 15 61.6 12.94

Total 54 49 18.51 84 38.4 17.26

Component 3: Rim and walls configuration

0. Smooth regular rim and no wall formation – – – – – –

1. Beginning walls with a thick, smooth regular rim 3 38 18.93 5 17.8 1.30

2. Visible walls that are thick and smooth with a scalloped or slightly wavy rim 25 44 17.05 25 24.1 3.55

3. Transitional stage 11 46 16.15 20 34.3 11.62

4. The rim sharper and increasingly irregular with more frequent bony projection.

The walls show further thinning.

10 63 16.34 16 49.5 11.21

5. Texture shows extreme friability and porosity. 5 62 17.92 16 58.2 11.53

Total 54 49 18.51 82 37.8 16.67

J.I. Cerezo-Roman, P.O. Hernandez Espinoza / Forensic Science International 236 (2014) 196.e1–196.e6196.e4

with the sternal end of rib bias and inaccuracy from the study ofSaunders et al. [8] (Table 8). Saunders et al. [8] presented a blindtest of four morphological methods of adult age-at-deathestimation using a sample from a 19th century Canadianpioneer cemetery. Among the methods that they used is thesternal end of the ribs, and they mention that the sample variedfrom 27 to 49 individuals. However, in their article the samplenumber is not presented and for this reason it is not presented in

Table 5Known age and component 1 pit depth.

ANOVA

Real chronological age

Sum of squares df

Between groups 1035.031 2

Within groups 17,130.302 51

Total 18,165.333 53

Table 6Known age and component 2 pit shape.

ANOVA

Real chronological age

Sum of squares df

Between groups 3010.428 4

Within groups 15,154.905 49

Total 18,165.333 53

Table 7Known age and component 3 rim and wall configuration.

ANOVA

Real chronological age

Sum of squares df

Between groups 4173.561 4

Within groups 13,991.772 49

Total 18,165.333 53

Table 8. Different for Saunders et al. [8] the sample in this studypresent under and over estimation of ages. However, there biaswas lower in the 17–29 and 40–49 age groups, while in thecurrent study they are slightly higher. In the study of Saunderset al. [8] there is an increase of bias in age categories older than50 years and an underestimation of ages in individuals olderthan 40 years, both of which are similar findings in the currentstudy.

Mean square F Sig.

517.516 1.541 0.224

335.888

Mean square F Sig.

752.607 2.433 0.06

309.284

Mean square F Sig.

1043.39 3.654 0.011

285.546

Fig. 1. Comparison of known age and estimate age for each male.

J.I. Cerezo-Roman, P.O. Hernandez Espinoza / Forensic Science International 236 (2014) 196.e1–196.e6 196.e5

4. Discussion

In this study, age changes related to the sternal end of the riband the method developed by Is can et al. [12] were examined on awell-documented sample of modern males from Mexico. This wasdone by observing the relationship between known age and themorphological changes in this area through descriptive statisticsand one-way statistical analysis of variance (one-way ANOVA).Using the method proposed by Is can et al. [12], we also evaluatedthe bias and inaccuracy between the known and estimated ages.

The analysis of descriptive statistics for each phase andcomponent suggest that as known ages increment the phases ofeach component also increment. However, there is not a cleardelimitation between the known age and component increments.Also, when the current results are compared with those from Is canet al. [12], the known age changes occurred at older ages in all thephases and components. The increment in ages in the study ofIs can et al. [12] are also lower than the results in our study. Also inthe study of Is can et al. [12] the standard deviations are also lowerthan the results in our study. Oettle and Steyn [37] also performeda similar study, applying similar procedures on a South Africasample. However, their results were different from both ourinvestigation and from the results of Is can et al. [12]. In theirsample, the changes that occurred with age occurred at earlierages. Yavuz et al. [36] also observed that the degenerativeprocesses were underestimated, particularly before 40 years ofage.

The second analysis performed was a one-way analysis ofvariance or one-factor ANOVA. The results from this statistical

Table 8Bias and inaccuracies for the Is can et al. sternal end of the rib estimates.

N Bias Inaccuracies

Known age this study

21–30 14 �6.8 8.2

31–40 12 �11.5 12.6

41–50 15 �9.3 10.4

51–60 7 �8.7 12.2

60+ 22 �12 18.7

Saunders et al. [8]

Sternal end of ribs known age

17–29 – 0.8 5.0

30–39 – 11.1 11.1

40–49 – �2.5 7.1

50–59 – �9.1 9.1

60+ – �15.5 16.6

analysis revealed that there was a statistically significantrelationship between the known age and the rim and wallconfiguration. The results we obtained are different from thosereported by Is can et al. [12]. The results of Is can et al. [12] suggestthat all the components had significant values, while we found thatrim and wall configurations were the most dependent on age in oursample. The one-factor ANOVA analysis reveals that differencesexist between known ages and rib components.

The analysis of bias and inaccuracy suggest that ages between21 and 30 present the lowest biases and inaccuracies. However,these values deviate dramatically after the age of 31, and after theage of 60 there is the highest increase. It was found that the methodunderestimates the known ages of individuals. When the bias andinaccuracy results are compared with other studies, such as theone by Saunders et al. [8], a similar pattern was observed in thegroup ages after 40 years. In the current research and in Saunderset al. [8] there is an underestimation of the ages, particularly true inindividuals older than 50 years. Most of the changes observed inthe sternal end of the rib are related to degenerative changes whichare difficult to interpret and likely show variation related to life-style, environment and activities [8,9]. Therefore, it is notsurprising that as the individual age biases and inaccuraciesassociated with age estimates also will increase.

This study confirms conclusions of previous studies thatsuggest there are variations between populations around theworld and changes in human remains related to age [9,45].Considering this, it is necessary to continue exploring morpholog-ical variation among populations and differential changes throughtime that occur with age to more accurate estimate age-at-deathand to have a broader understanding of human variation.

5. Conclusions

Most research on estimating age-at-death using humanskeleton material suggested that degenerative processes are goodindicators for age estimates. However, these processes can varyamong populations, depending upon factors such as biologicalaffinity and relate to growth and development, as well as life-conditions and lifestyle. The majority of indicators and techniquesused to determine age-at-death by physical anthropologists in theUnited States are derived from individuals who are genetically andmorphologically different from Mexican populations.

The objective of this research was to evaluate the precision ofthe method proposed by Is can et al. [12] to estimate the age-at-death using the sternal end of the rib on a Mexican sample. Weselected specimens from two groups of known age and sex. It was

J.I. Cerezo-Roman, P.O. Hernandez Espinoza / Forensic Science International 236 (2014) 196.e1–196.e6196.e6

possible to observe the relationship between morphologicalchanges and known age. In the case of the fourth rib, pit formand rim and wall configurations had a significant relationshipswith age. Once we knew the characteristics of the ribs, the age-at-death of individuals in our sample was determined using thestandards of Is can et al. [12], without knowing their known ages.The results suggest that ages were underestimated. This wasparticularly evident in individuals older than 40 at the time ofdeath. Understanding the relationships with age changes throughtime are diagnostic and essential to a more precise determinationof the age-at-death.

The results of the analysis of the timing and forms ofdegenerative processes in the sample population used in thisstudy differ from the results of previous research using the sternalend of the rib [12,35,36,41,42]. These new results suggest thatdegenerative processes occurring with age in this Mexicanpopulation occurred later than in other populations. We hopethat these results will promote the evaluation of methods forestimation of age-at-death and a more critical consideration offactors that influence human variation.

Acknowledgments

We would like to thank Dr. Jose Ramon Fernandez Caceres,Mirna Martınez Garcıa M.D., Sergio Ubando Lopez M.D., LuisPacheco, Roberto Medina, Daniel Trejo Lopez, and FabiolaGuadalupe Gutierrez Sanchez M.D. from the Servicio MedicoForense de la Ciudad de Mexico for providing access to thematerial. We would also like to thank Arturo Talavera for giving usthe opportunity to study the collection of San Nicolas Tolentino.We are indebted to Dr. James Lu Zhenqiang (University of Arizona)for statistical advice. Dr. Marıa Eugenia Pena (Escuela Nacional deAntropologıa e Historia), Dr. Lourdes Marquez Morfin (EscuelaNacional de Antropologıa e Historia), and Aldo J. Martınez providedsupport throughout the research. We would like to thank Dr.Thomas Fenn (Katholieke Universiteit Leuven) and Megan Sheehan(University of Arizona) for their help in editing this article. Also, weare very grateful to Dr. John McClelland (University of Arizona),Prof. Wim Van Neer (Royal Belgian Institute of Natural Sciences),Bruce Anderson (University of Arizona), and Jane Buikstra (ArizonaState University) who assisted immeasurably with their feedbackand support. All errors are the responsibility of the authors.

References

[1] M. Lampl, M.L. Johnson, Problems in the aging of skeletal juveniles: perspectivesfrom maturation assessments of living children, Am. J. Phys. Anthropol. 101(1996) 345–355.

[2] T.D. White, P.A. Folkens, The Human Bone Manual, Elsevier-Academic Press,Burlington, MA, 2005.

[3] M.Y. Is can, K.A.R. Kennedy, Reconstruction of Life from the Skeleton, John & Sons,Inc., New York, 1989.

[4] D.A. Komar, J.E. Buikstra, Forensic Anthropology Contemporary Theory andPractice, University Press, New York, 2008.

[5] J.E. Buikstra, D. Ubelaker, Standards for Data Collection from Human SkeletalRemains: Proceeding of a Seminar at the Field Museum of Natural HistoryOrganized by Jonathan Haas, Arkansas Archaeology Survey, Fayetteville, 1994.

[6] M. Cox, Aging adults from the skeleton, in: M. Cox, S. Mays (Eds.), HumanOsteology: In Archaeology and Forensic Science, Greenwich Medical Media,London, 2000.

[7] E.H. Kimmerle, D.A. Prince, G.E. Berg, Inter-observer variation in methodologiesinvolving the pubic symphysis, sternal ribs, and teeth, J. Forensic Sci. 53 (2008)594–600.

[8] S.R. Saunders, C. Fitzgerald, R.C. Dudar, H. Mc Killop, A test of several methods ofskeletal age estimation using a documented archeological sample, Can. Soc.Forensic Sci. 25 (1992) 97–118.

[9] S.M. Hens, E. Rastelli, G. Belcastro, Age estimation from the human os coxa: a teston a documented Italian collection, J. Forensic Sci. 53 (2008) 1040–1043.

[10] L. Fanton, M.-P. Gustin, U. Paultre, B. Schrag, D. Malicier, Critical study ofobservation of the sternal end of the right 4th rib, J. Forensic Sci. 55 (2010)467–472.

[11] D. Ubelaker, Forensic anthropology: methodology and diversity of applications,in: A.M. Katzenberg, S.R. Saunders (Eds.), Biological Anthropology of the HumanSkeleton, 2nd ed., John Wiley & Sons, Inc., Hobeken, NJ, 2008, pp. 41–69.

[12] M.Y. Is can, S.R. Loth, R.K. Wright, Metamorphosis at the sternal rib end: a newmethod to estimate age at death in white males, Am. J. Phys. Anthropol. 65 (1984)147–156.

[13] M.Y. Is can, S.R. Loth, R.K. Wright, Biological variation in sternal extremity of therib and its effect on age determination, J. Forensic Sci. 32 (1987) 452–466.

[14] M.Y. Is can, S.R. Loth, Determination of age from sternal rib by the phase analysis inwhite females: a test of the phase method, J. Forensic Sci. 31 (1986) 90–92.

[15] K. Nishino, Studies on the human rib-cartilage, Kekkaku 44 (1969) 131–137.[16] A.A. Semine, A. Damon, Costochondral ossification and aging in five populations,

Hum. Biol. 47 (1975) 101–116.[17] W.F. McCormick, Mineralization of the costal cartilages as an indicator of age:

preliminary observation, J. Forensic Sci. 25 (1980) 736–741.[18] W.F. McCormick, J.H. Stewart, Age related changes in the human plastron: a

roentgenographic and morphologic study, J. Forensic Sci. 33 (1988) 100–120.[19] F. Dedouit, S. Bindel, D. Gainza, A. Blanc, F. Joffre, D. Rouge, N. Telmon, Application

of the Iscan method to two- and three-dimensional imaging of the sternal end ofthe right fourth rib, J. Forensic Sci. 53 (2008) 288–295.

[20] E.D. Sedlin, H.M. Frost, A.R. Villanueva, Variation in cross-section area of rip cortexwith age, J. Gerontol. 18 (1963) 9–13.

[21] B.N. Epker, M. Kelin, H.M. Frost, Magnitude and location of cortical bone loss inhuman rip with aging, Clin. Orthop. 41 (1965) 198–203.

[22] H.M. Frost, Supracellular organization of bone turnover (remodeling) in largemammals and its parameters, in: Proceedings of the First Workshop on BoneMorphometry, Ottawa, 1976.

[23] D.M. Mulhern, Rib remodeling dynamics in a skeletal population from Kulubnarti,Nubia, Am. J. Phys. Anthropol. 111 (2000) 519–530.

[24] Y.-S. Kim, D.-I. Kim, D.-K. Park, J.-H. Lee, N.-E. Chung, W.-T. Lee, S.-H. Han,Assessment of histomorphological features of the sternal end of the fourth ribfor age estimation in Koreans, J. Forensic Sci. 52 (2007) 1237–1242.

[25] E.R. Kerley, Estimation of skeletal age: after about age 30 years, in: T.D. Stewart(Ed.), Personal Identification in Mass Disasters, National Museum of NaturalHistory, Washington, DC, 1970, pp. 57–70.

[26] D.H. Ubelaker, Human Skeletal Remains: Excavation, Analysis, Interpretation,Smithsonian Institution, Taraxacum, Washington, 1978.

[27] E. Rist, L. Gally, C. Trocme, L’ossification des cartilages costaux dans I’especehumain, Presse Medicale 41 (1928) 641–644.

[28] L. Lichtenstein, Diseases of Bone and Joints, C.V. Mosby, Saint Louis, 1975.[29] J.C. Dudar, An evaluation of intercostal variation in adult sternal rib ends, Can. Soc.

Forensic Sci. Program 23 (1992) 139.[30] K.M. Hartnett, Analysis of age-at-death estimation using data from a new, modern

autopsy sample—Part II: Sternal end of the fourth rib, J. Forensic Sci. 55 (2010)1151–1156.

[31] T.A. Rathbun, Heads or tails? Gender and age indicators of the cranial andinfracranial skeleton, Am. Acad. Forensic Sci. Program 107 (1988).

[32] C.W. Cargill, J.M. Suchey, Value of the sternal rib end in the determination of age inunidentified bodies, Am. Acad. Forensic Sci. Program (1987).

[33] M.Y. Is can, S.R. Loth, R.K. Wright, Age estimation from the rib by phase analysis:white female, J. Forensic Sci. 30 (1985) 41–44.

[34] S.R. Loth, Age assessment of the spitalfields cemetery population based on ribphase analysis, Am. J. Phys. Anthropol. 16 (1993) 135.

[35] M.Y. Is can, The aging process in the rib: an analysis of sex and race relatedmorphological variation, Am. J. Hum. Biol. 3 (1991) 617–623.

[36] F.M. Yavuz, Y.M. Is can, S.A. Cologlu, Age assessment by rib phase analysis in Turks,Forensic Sci. Int. 98 (1998) 47–54.

[37] A.C. Oettle, M. Steyn, Age estimation from sternal ends of ribs by phase analysis inSouth African blacks, J. Forensic Sci. 45 (2000) 1071–1079.

[38] M.C. Meena, Y. Rani, M. Rani, Bilateral metamorphological variation at sternal endof fourth rib, Euras. J. Anthropol. 3 (2012) 41–46.

[39] C. Masset, Ou en est la paleodemographie? Bull. et Mem. de la Soc. Anthropol.Paris 2 (1990) 109–122.

[40] M.Y. Is can, S.R. Loth, Osteological manifestations of age in adults, in: M.Y. Is can,A.R. Kennedy (Eds.), The Reconstruction of Life from Skeleton, Alan R. Lass, Inc.,New York, 1986.

[41] M.Y. Is can, Age Markers in the Human Skeleton, Charles C. Thomas, Springfield, IL,1989.

[42] C. Yoder, D. Ubelaker, Examination of variation in sternal rib end morphologyrelevant to age assessment, J. Forensic Sci. 46 (2001) 223–227.

[43] E. Nikita, Quantitative assessment of the sternal rib end morphology andimplications for its application in aging human remains, J. Forensic Sci. 58(2013) 324–329.

[44] A.L. Santos, How old is this pelvis? A comparison of age at death estimation usingthe auricular surface of the ilium and os pubis, in: Aspects of African Archaeology:Papers for the 10th Congress of the PanAfrican Association for Prehistory andRelated Studies, 1996, 29–36.

[45] A. Schmitt, P. Murail, E. Cunha, D. Rouge, Variability of the pattern of aging in thehuman skeleton: evidence from bone indicators and implication on age at deathestimation, J. Forensic Sci. 47 (2002) 1203–1209.