ORIGINAL ARTICLE

The persistence of epiphyseal scars in the adult tibia

Catriona Davies & Lucina Hackman & Sue Black

Received: 6 December 2012 /Accepted: 20 February 2013# Springer-Verlag Berlin Heidelberg 2013

Abstract Estimation of chronological age from skeletalmaterial is dependent upon estimation of maturational stageobserved. Following completion of epiphyseal fusion, atransverse radio-opaque line, termed “epiphyseal scar”,may be observed in the region of the former growth plate.According to the literature, this line is likely to becomeobliterated shortly after completion of epiphyseal fusion.Consequently, presence of an epiphyseal scar has beeninterpreted as an indication of recent epiphyseal fusion;however, this has not been validated by quantitative re-search. A study was undertaken to determine persistenceof the epiphyseal scars in a cross-sectional population ofadults between 20 and 50 years of age. This study examined1,216 radiographs of proximal and distal tibiae from bothsexes and sides of the body. This study suggested that98.05 % of females and 97.74 % of males retained someremnant of the epiphyseal scar at the proximal tibia whilst92.72 % of females and 92.95 % of males retained someremnant of the epiphyseal scar at the distal tibia. Generallinear model (GLM) analysis determined that chronologicalage accounted for 2.7 % and 7.6 % of variation in persis-tence of the epiphyseal scar at the proximal and distal tibiae,respectively. This study suggests that obliteration of theepiphyseal scar is not as dependent on chronological ageas previously thought. It is, therefore, recommended thatthis feature not be used as an indicator of chronologicalage during forensic age assessment.

Keywords Forensic anthropology . Epiphyseal scar . Ageestimation . Tibia . Radiographs

Introduction

Estimation of chronological age is a necessary step in es-tablishment of a biological profile in forensic human iden-tification. Formed from data relating to characteristics ofage, sex, stature and ancestral origin, the biological profilefacilitates a targeted search of missing persons' records in anattempt to assign identity.

Age estimation relates biological changes within hardtissues of the body to the chronological age of an individual.Although many approaches to age estimation exist [1–9],the method applied is dependent upon the anatomical re-gions of the deceased recovered, the approximate stage ofdevelopment of the individual (i.e., juvenile or adult) andthe manner in which the remains are presented for exami-nation, e.g., physical remains or medical imaging [10].Assessments of age can be based on an evaluation of skel-etal morphology through physical examination of the re-mains or from images resulting from medical imagingincluding radiography or computed tomography (CT)[11–16]. Application of medical imaging to age estimationallows an assessment of age to be based on more thanexternal morphology of bone and may highlight age-related changes to the underlying trabecular structure [17]and density [18].

The epiphyseal scar is one of the characteristics that maybe observed through radiographic imaging [19]. Althoughremaining visible on cross section and on radiographs, onceepiphyseal fusion is complete, it is not possible to observethe epiphyseal scar on the external surface of the bone [20].Although their presence is noted in radiographic methods ofage estimation such as the atlases by Greulich and Pyle [2],Pyle and Hoerr [3] and Hoerr et al. [21], it is generallyreported that epiphyseal scars will become obliterated overtime [22–24]. Observation of a clear line has beeninterpreted as an indication of recent epiphyseal fusion

C. Davies : L. Hackman : S. Black (*)Centre for Anatomy and Human Identification, College of LifeSciences, University of Dundee, Dow Street,Dundee DD1 5EH, UKe-mail: s.m.black@dundee.ac.uk

Int J Legal MedDOI 10.1007/s00414-013-0838-3

[23] and consequently, the obliteration of the epiphysealscar is often used as a biological criterion in methods ofage estimation based on attributing numerical scores tomaturity stages [14, 25, 26].

Persistence of the epiphyseal scar in adults was firstsuggested by Cope in 1920 who also highlighted the riskof misinterpretation of this line as an indicator of recentfusion [19]. In his study of epiphyseal fusion lines, particu-lar attention was drawn to the persistence of lines at thedistal femur and proximal tibia, where a thin line of greaterrelative density was regularly observed.

In their radiographic study of the growing knee, Pyle andHoerr [3] noted that although the epiphyseal line was clearlyobservable distal to the lateral tibial condyle in the finalmaturity stage, obliteration had commenced in the areadistal to the medial condyle; however, the authors did notattempt to attach a time frame to the obliteration of theepiphyseal scar. The radiographic study of the developingfoot and ankle by Hoerr et al. [21] noted the presence of theepiphyseal scar in adult individuals at the distal tibia; how-ever, contrary to the study of the proximal tibia by Pyle andHoerr [3], the Hoerr et al. atlas [21] suggested that theepiphyseal scar or terminal line may remain visible through-out life.

More recently, Weiss et al. [27] have advised caution tothose using the presence of an epiphyseal scar as an indica-tor of a recently fused epiphysis. In their study of 131 firstmetatarsal bones from individuals aged between 17 and88 years, epiphyseal scars were present at the epiphysis ofthe base of the first metatarsal in 38 % of their sample.During the course of their study, no statistical relationshipwas found between chronological age and the obliteration ofthis epiphyseal scar. This appears to be one of the firststudies to attempt quantification of the persistence of epiph-yseal scars. It also appears to be one of the first studies tosuggest that the assumption on which the final stage ofmany methods of age estimation, which rely on the obliter-ation of the epiphyseal scar, are based may be erroneouswhen applied to certain bones.

In their study regarding the time frame for ossifica-tion of the distal radial and ulnar epiphyses, Baumannet al. [15] observed that the youngest individuals toexhibit obliteration of the epiphyseal scar at the distalradius were 16.2 and 18.7 years in females and males,respectively; however, an epiphyseal scar was present atthis region in some individuals up to 30.8 and 31.0 yearsin females and males, respectively. Baumann et al. [15]indicate that their results do not exclude the possibilitythat the epiphyseal scar of the distal radius may beretained into adulthood in some individuals. This, inturn, raises a more general question regarding validityof the inclusion of the epiphyseal scar as an age-relatedmaturity criterion in other anatomical regions.

Epiphyseal scars have been noted in the clinical litera-ture, in particular, in relation to internal fixation of proximaland distal femoral fractures [28, 29]. Within these studies,the epiphyseal scars of the proximal and distal femoralepiphyses are frequently mentioned as anatomical land-marks from which the appropriate placement of an intra-medullary nail may be judged [30, 31]. It would, therefore,be reasonable to suggest that the presence of epiphysealscars at these anatomical sites in adult individuals is suffi-ciently consistent to merit their use as a landmark in surgery.Although few in number, the studies that exist within thebody of the literature suggest that there are unansweredquestions surrounding the observation of the epiphyseal scarand its association with chronological age.

There is a paucity of research studies concerning thepotential persistence of epiphyseal scars, and as a conse-quence, the lack of evidence of persistent epiphyseal scarswithin the literature may be incorrectly interpreted as evi-dence of their obliteration in adult individuals. The recentfindings of Weiss et al. [27] suggest that it is appropriate toquantify the persistence rate of epiphyseal scars at variousskeletal regions in an attempt to further the collectiveknowledge of epiphyseal scars and their value in ageestimation.

Materials and methods

A sample of 1,216 radiographs consisting of 617 from theproximal tibia and 599 from the distal tibia was collectedfrom separate individuals at Ninewells Hospital, Dundeeand all imaging studies had been conducted between 2008and 2011. Images included in this study were taken fromboth sides of the body and were obtained during clinicalassessment of an injury and taken in the anterior–posteriorplane. Individuals who were noted as having experienced adelay in skeletal development, an injury or chronic illnessthat may have affected the region of the growth plate or theepiphyseal scar were excluded from the sample. All radio-graphs were obtained for the purposes of clinical investiga-tion and were subsequently reported to be normal by aconsultant clinical radiologist.

For each image included in the sample, the side of thebody, date of birth of the individual (DoB) and date onwhich the image was taken (DoI) were recorded, fromwhich the chronological age of each individual was calcu-lated. Individuals included in the sample were between theages of 20 and 50 years at the time the radiograph was takenand were represented by a single radiograph. Ethical ap-proval for the use of anonymous clinical data in this studywas given by Tayside NHS Trust. The distribution of thesample according to sex, anatomical region and chronolog-ical age is presented in Table 1. According to the Scottish

Int J Legal Med

Government [32], approximately 1.9 % of the population ofTayside is non-White and there is approximately 16–19 %relative deprivation across the three counties included in theNHS Tayside population. It was not possible to accessinformation regarding ancestral groupings or socio-economic status of the sample population; however, as theradiographs represent a cross-section of the Tayside popu-lation, there is no reason to suggest that the clinical popula-tion would differ significantly.

Using Adobe Photoshop®, each image was divided intosix equally spaced tracks spanning from the medial to lateralextremities of the bone for both the proximal (Fig. 1) anddistal (Fig. 2) tibiae. A score was then assigned to the levelof persistence of the epiphyseal scar within each trackaccording to the degree of preservation of the epiphysealscar. A score of 0 was assigned if no scar was observed inthe track, a score of 1 was assigned if a partial or fenestratedscar was observed and a score of 2 was assigned if theepiphyseal scar was complete across the width of the track.The total persistence score (TPS) was then calculated as thesum of the scores assigned to each of the six tracks, i.e., amaximum score of 12 would indicate a solid epiphyseal scarspanning the entirety of the width of the bone and a mini-mum score of zero would indicate complete absence of thescar.

The TPS assigned to each individual was recorded andthen analysed using IBM SPSS™ statistics software.

Relationships between chronological age, sex, side of thebody, end of the bone and TPS were assessed using a seriesof general linear model (GLM) analyses.

Intra-observer analysis

A subset of radiographs of 30 females and 30 males wasrandomly selected for both the proximal and distal tibiae.These images were re-assessed to determine variancebetween repeated observations by a single observer. Con-sistency between the repeated assessments was deter-mined by calculating the percentage of second roundscores that were within two of those assigned duringthe first round of assessment. This shall be referred toas the percentage agreement and was calculated for fe-males and males. In addition, a GLM analysis was car-ried out to determine the statistical significance of thevariance between TPS assigned on the first and secondrounds of assessment.

Inter-observer analysis

The subset of images used in the intra-observer study wasassessed by three observers to determine consistency betweenTPS assigned by multiple observers. All three observers usedin this study hold a Ph.D. in anatomy or forensic anthropolo-gy; however, their experience in radiographic interpretation ofage was variable. Observer 1, though not experienced in ageestimation or radiograph interpretation, has extensive experi-ence in line and pattern recognition; observer 2 is a highlyexperienced forensic anthropologist who specialises in foren-sic age estimation; and observer 3 is a highly experiencedforensic anthropologist.

The percentage agreement between observers was calcu-lated and a GLM analysis was carried out to determine thesignificance of the relationship between observer and TPS. To

Table 1 Sample distribution according to anatomical region, sex andchronological age

Age cohort(years)

Femaleproximaltibia

Maleproximaltibia

Femaledistaltibia

Maledistaltibia

20–29 100 100 99 98

30–49 98 100 97 90

40–50 110 109 106 109

Total 308 309 302 297

Fig. 1 Example of tracks applied to each proximal tibial image duringassessment

Fig. 2 Example of tracks applied to each distal tibial image duringassessment

Int J Legal Med

accurately determine variation in TPS between observers, thevariation in TPS that was attributable to biological sex wasalso considered.

Results

Intra-observer and inter-observer analysis

The greatest intra-observer percentage agreement (±2scores) was observed in the male proximal tibial assessmentgroup where 83.3 % of assessments were within two scoresof those assigned at the first attempt. Both male distal tibialand female proximal tibial agreement was rated at 80 %,whilst the lowest percentage agreement of 66.67 % wasfound at the female distal tibia. The results of the GLManalyses carried out on data derived from the intra-observerassessments are presented in Table 2. These analyses deter-mined that there was no statistically significant difference inTPS within the intra-observer data at either the proximal(P=0.658) or distal (P=0.192) tibia after variation due tobiological sex having been taken into account. The results ofthe intra-observer analyses, therefore, suggest that the meth-od is reliable when applied by a single observer on multipleoccasions.

The percentage agreement between observers was calcu-lated and the results presented in Table 3. These resultsshowed that the greatest percentage agreement (100 %)was found between observers one and two in the femaleproximal tibial sample. The lowest percentage agreement(66.67 %) was observed between observers two and three atthe male proximal tibia. The results of the GLM analysescarried out on the data derived from the inter-observerassessments are presented in Table 4. These resultssuggested that once the influence of variation in sex onTPS had been accounted for, the variation observed inTPS assigned by multiple observers was not statisticallysignificant at either the proximal (P=0.618) or distal (P=

0.616) tibia. The method can, therefore, be consideredrepeatable.

Main analysis

Distributions of TPS amongst the samples are presented inFigs. 3 and 4 for the proximal and distal tibiae, respectively.Overall, 98.05 % of females and 97.74 % of males wereobserved to exhibit some remnants of an epiphyseal scar atthe proximal tibia. Initial analysis showed a statistical dif-ference to exist between sexes (P<0.001); consequently,biological sex was included as an explanatory variable inall subsequent analyses. A Shapiro–Wilk normality testshowed that neither the female nor male data sets werenormally distributed (W-statistic=0.977; P=0.001).

Overall, 92.72 % of females and 92.95 % of maleswere observed to exhibit a degree of retention of theepiphyseal scar at the distal tibia. A Shapiro–Wilk nor-mality test showed that neither the female (W-statistic=0.971; P<0.001) nor the male (W-statistic=0.968; P<0.001) samples were normally distributed according toTPS. A one-way ANOVA suggested that a statisticallysignificant difference existed between the TPS assignedto females and males (P=0.012). As with the proximaltibia, biological sex was included as an explanatoryvariable in all subsequent analyses.

Mean chronological age of the individuals to whom eachpersistence score was assigned was calculated. These resultsare presented in Figs. 5 and 6 for proximal and distal tibiae,respectively. Organisation of the data in this way revealedthe presence of an inverse relationship between mean chro-nological age and TPS at both anatomical sites. To deter-mine the statistical significance of the relationship betweenchronological age and TPS, GLM analyses were carried out.This also facilitated quantification of the effects of sex andside of the body on TPS and, therefore, the persistence ofthe epiphyseal scar.

Results of these analyses are presented in Tables 5 and 6for the proximal and distal tibiae, respectively. These datasuggest that when considered as independent variables, bi-ological sex, side of the body and chronological age exhibitstatistically significant relationships with TPS. When theinteractions between these variables were examined at theproximal tibia, only the relationship between chronological

Table 2 Results of intra-observer general linear model analyses

Region P value R Adjusted R2

Proximal tibia 0.658 0.19 0.011

Distal tibia 0.192 0.24 0.026

Table 3 Inter-observer percent-age agreement according to ana-tomical region and sex

Region Obs 1 versus Obs 2 Obs 1 versus Obs3 Obs 2 versus Obs3

Male distal tibia 73.33 83.33 83.34

Female distal tibia 86.66 83.33 90.00

Male proximal tibia 83.33 80.00 66.66

Female proximal tibia 100.00 86.66 86.67

Int J Legal Med

age, sex and TPS was statistically significant (P=0.03). Asthis combination displayed the highest statistical signifi-cance and the greatest coefficient of determination (R2=0.101), this combination explains the greatest variationwithin TPS data set at the proximal tibia.

In contrast, the GLM analysis carried out on the dataderived from the analysis of the distal tibia suggestedthat the factors of biological sex and side of the bodyprovided the best model for explaining variation in TPSat this anatomical site (P=0.001; (R2=0.27). To deter-mine the variation in persistence of the epiphyseal scar atdifferent areas of the bone, the persistence score for eachof the medial, central and lateral regions was calculatedas the sum of the persistence scores within the respectivetracks (tracks 1 and 2=medial, tracks 3 and 4=centraland tracks 5 and 6=lateral as shown in Fig. 1). Generallinear model analyses were applied to the data from boththe proximal and distal tibiae to determine the statisticalsignificance of the relationship between region and TPSwhilst considering the effects of previously discussedfactors. This analysis suggested that once the variationdue to biological sex, side of the body and chronologicalage had been taken into consideration, the relationshipbetween region and TPS was not statistically significantat either the proximal (P=0.933) or distal (P=0.660)tibia. These results suggest that variation in the persis-tence of the epiphyseal scar within the medial, centraland lateral regions of the proximal and distal tibiae is notrelated to their position within the bone.

Proximal tibia versus distal tibia

To determine the statistical significance of the variance inTPS at the two anatomical locations and therefore assess thepotential variation in persistence of the epiphyseal scarwithin a single bone, a GLM analysis was carried out. Allpreviously examined variables were included in this analy-sis as potential explanatory factors. The results of this anal-ysis are presented in Table 7.

This analysis suggests that the variation in TPS betweenthe proximal and distal tibiae was not statistically significant(P=0.789, R2=−0.001). Subsequent analyses were under-taken to determine the significance of the combined influ-ence of chronological age, biological sex, side of the bodyand end of the bone on the TPS. This analysis suggested thatthe relationship between TPS and the combined biologicalfactors included in this study is not statistically significant(P=0.081; R2=0.159); however, the combined factors doexplain 15.9 % of variation in TPS.

Table 4 Results of inter-observer GLM analysis

Region P value R Adjusted R2

Proximal tibia 0.618 0.217 0.019

Distal tibia 0.616 0.179 0.014

0

5

10

15

20

25

30

35

40

45

50

0 1 2 3 4 5 6 7 8 9 10 11 12

Nu

mb

er o

f In

div

idu

als

(n=5

99)

Total Persistence Score

Females

Males

Fig. 4 Sample distribution according to sex and total persistence scoreat the distal tibia

0

10

20

30

40

50

60

70

0 1 2 3 4 5 6 7 8 9 10 11 12

Nu

mb

er o

f In

div

idu

als

(n=6

17)

Total Persistence Score

Females

Males

Fig. 3 Sample distribution according to sex and total persistence scoreat the proximal tibia

0

5

10

15

20

25

30

35

40

45

50

0 1 2 3 4 5 6 7 8 9 10 11 12

Mea

n C

hro

no

log

ical

Ag

e (y

ears

)

Total Persistence Score

Males

Females

Fig. 5 Mean chronological age of each persistence score at the prox-imal tibia according to sex

Int J Legal Med

Discussion

Medical imaging, and in particular radiographic imaging, isa valuable resource in the post-mortem examination of hu-man remains. Through the analysis of radiographic images,it is possible to assess the level of skeletal maturity of theindividual and consequently estimate their chronologicalage [2, 3, 14, 21, 33, 34]. The methods used to determinechronological age from the degree of skeletal maturity relyon the progressive ossification of skeletal elements, culmi-nating in the fusion of the epiphyses to their respectivediaphyses [2, 3]. Following completion of epiphyseal fu-sion, a transverse radio-opaque line may be observed in theregion of the former growth plate.

Persistence of transverse radio-opaque lines andbands at various anatomical regions in adult individualshas been noted in the archaeological [35, 36], forensic[27] and clinical [30] literature. Their utility in forensicage estimation, however, has remained largely un-challenged despite questions being raised as to theirvalue. A recent article by Weiss et al. [27] appears tobe the first report that specifically questions the assump-tions regarding epiphyseal scars and their application inforensic age estimation.

In contrast to the previously held belief that persistenceof an epiphyseal scar in adults would only occur in aminority of the population, this study found that epiphysealscars persisted at the proximal tibia in over 98 % of femalesand 97 % of males between the ages of 20 and 50 years. Thetotal persistence rate at the distal tibia was observed to belower than that at the proximal tibia; however, epiphysealscars were still observed in over 92 % of both female andmale individuals.

Although chronological age displayed a significant rela-tionship with TPS when considered as an independent fac-tor, statistical analysis of the data obtained from theproximal and distal tibiae suggested that chronological agewas only responsible for 2.7 % and 7.6 % of variation inTPS at the proximal and distal tibiae, respectively. The re-sults of this study suggest that chronological age does notplay as significant a role in determining the degree ofpersistence of the epiphyseal scar at either the proximal ordistal tibia as was previously thought. This study also sug-gests that factors other than chronological age may be re-sponsible for the majority of variation in persistence of theepiphyseal scar.

The skeleton is a complex system that must balancestructural function and metabolic requirements of thebody, characteristics which differ according to the sexand the life stage of the individual. Although the struc-tural function of the skeleton is consistent betweenfemales and males, there is greater interaction betweenmetabolic activity and the skeleton in females thanmales due to the potential calcium requirements of thereproductive system [37].

Similar distribution patterns were observed in both sexsamples; however, an analysis of the data suggested therewas a significant difference between persistence of epiphy-seal scars in females and males at both anatomical locationsconsidered in this study. The overall persistence rate at theproximal tibia indicates that females are more likely toexhibit an epiphyseal scar in this region than males whilstthe converse is true at the distal tibia where the overallpersistence rate was marginally greater in males thanfemales.

0

5

10

15

20

25

30

35

40

45

0 1 2 3 4 5 6 7 8 9 10 11 12Mea

n C

hro

no

log

ical

Ag

e (y

ears

)

Total Persistence Score

Male

Female

Fig. 6 Mean chronological age of each persistence score at the distaltibia according to sex

Table 5 Results of the general linear model of the proximal tibia

Factor(s) Significance R2 Adjusted R2

Age 0.029 0.074 0.027

Sex <0.001 0.047 0.045

Side <0.001 0.037 0.036

Age * sex 0.03 0.19 0.101

Age * side 0.555 0.156 0.063

Sex * side 0.532 0.084 0.08

Age * sex * side 0.384 0.313 0.147

Table 6 Results of the general linear model of the distal tibia

Factors Significance R2 Adjusted R2

Age <0.001 0.123 0.076

Sex 0.012 0.011 0.009

Side 0.146 0.004 0.002

Age * sex 0.01 0.21 0.12

Age * side 0.137 0.184 0.092

Sex* side 0.001 0.32 0.27

Age * sex * side 0.204 0.337 0.166

Int J Legal Med

Although biological sex displayed a significant relation-ship with TPS at both anatomical regions, the data from thisstudy suggest that the influence of sex is greater at theproximal tibial epiphyseal scar than the distal tibial epiphy-seal scar. It should be concluded, therefore, that althoughstatistically significant variation does exist in the persistenceof epiphyseal scars in females and males, biological sexonly explains a small percentage of the total variation inTPS.

The third factor considered by this study was the influ-ence of side of the body on the persistence of the epiphysealscar. It is currently reported that approximately 90 % ofhumans preferentially use their right hand in movementsor actions that require the use of a single hand [38–40]. Thisfunctional bias towards one side may result in bilateralasymmetry in bone formation and resorption due to in-creased muscle tone of the dominant side within the limbconcerned [41]. In contrast to the upper limb where domi-nance of the right side is exhibited, the left side is often thedominant side in the lower limb in a pattern termed “crossedsymmetry” [41, 42]. Despite this dominance, the potentialfor asymmetry in the lower limb is lower than in the upperlimb due to functional constraints [41]. Although this theoryhas not, to the authors' knowledge, been applied to thepersistence of transverse radio-opaque lines, it has beenused to explain enthesopathies or occupational stressmarkers in archaeological populations [41, 43] .

The results of this study suggested that there was nosignificant relationship between side of the body and TPSat the distal tibia (P=0.146); however, a significant interac-tion was observed at the proximal tibia (P=<0.001) where3.6 % of total sample variation was attributable to side of thebody when considered as an independent factor. The re-lationships between side of the body and either biological

sex or chronological age and TPS was statistically insignif-icant (P=0.532 and P=0.555, respectively). These resultsindicate that variation explained by side of the body couldalso be explained by biological sex or chronological age.These results, therefore, suggest that side of the body doesnot play a significant role in determining the persistence ofthe epiphyseal scars of the tibia within this study population.

When the proximal tibia and distal tibia were compared,fewer individuals were observed to display some remnantsof the epiphyseal scar at the distal tibia than at the proximaltibia. There is, therefore, a suggestion that chronological ageand remodelling influence may exert a greater influence onthe persistence of the epiphyseal scar at the distal tibiacompared with the proximal tibia. It is recorded in theliterature that there is a bias in the persistence of Harris linesat the distal tibia compared with other regions of the skele-ton [35, 36]. To determine whether this bias was also presentin the persistence of epiphyseal scars, a comparative analy-sis was carried out to establish the variance in TPS at theproximal and distal tibiae.

The results of this study indicate that the persistence ofthe epiphyseal scar within the tibia was not significantlyinfluenced by the end of the bone (P=0.789). This suggeststhat the formation and remodelling of epiphyseal scars maybe different to other transverse radio-opaque bands [44].

Although no significant difference was found in TPSbetween the proximal and distal ends of the tibia, the studyis limited by the fact that the radiographs used in the anal-ysis of the proximal and distal tibiae, respectively, werefrom a cross section of the population. Therefore, althoughthe conclusions drawn from this data are statistically validand may represent the population as a whole, there has beenno opportunity to study the TPS of different skeletal siteswithin one individual.

Table 7 General linear model oftotal persistence score at theproximal tibia compared withdistal tibia

Factor(s) Significance R R2 Adjusted R2

Age <0.001 0.27 0.073 0.049

Sex <0.001 0.15 0.023 0.022

Side 0.014 0.07 0.005 0.004

End 0.789 0.03 <0.001 −0.001

Age * sex 0.003 0.37 0.138 0.092

Age * side 0.331 0.32 0.103 0.056

Age * end 0.017 0.33 0.11 0.064

Sex * side 0.025 0.18 0.032 0.03

Sex * end 0.209 0.15 0.024 0.022

Side * end 0.001 0.12 0.015 0.012

Age * sex * end 0.151 0.45 0.205 0.115

Sex * side * end 0.116 0.21 0.045 0.039

Age * side * end 0.215 0.42 0.178 0.085

Age * sex * side 0.397 0.44 0.195 0.105

Age * sex * side * end 0.081 0.57 0.329 0.159

Int J Legal Med

It should be acknowledged that there is a potential forconfusion between Harris lines and epiphyseal scars. Al-though Harris lines were noted within the diaphyses of someindividuals within this study sample, on no occasion was anepiphyseal scar interpreted as a Harris line or vice versa. AsHarris lines are believed to form during the recommencementof growth following a temporary arrest, it is unlikely that aHarris line would form in the exact position of the epiphysealscar and it is, therefore, improbable that an experienced ob-server would fail to differentiate the two structures.

In addition to the implications for forensic age assess-ment in deceased individuals, the results of this study mayimpact methodologies employed in methods of age assess-ment in living individuals, where the disappearance of theepiphyseal scar is commonly applied as the final maturityindicator [45–48]. It should be acknowledged, however, thatthe majority of these studies concern the medial clavicle[48–50], an area in which the epiphyseal scar does becomeobliterated due to the type of epiphysis formed and thestresses to which the bone is exposed by the surroundinganatomy. It must also be acknowledged that these studies areconcerned with images obtained from CT and not plain filmradiography. The detection and interpretation of epiphysealscars on CT images are dependent upon a number of factorsincluding slice thickness and resolution of the imagingsystem on which the scans were taken [49]. Consequently,the findings from studies based on conventional radiographysuch as this study and that by Weiss et al. [27] may not bedirectly applicable to images obtained through other imag-ing modalities. This, however, should form the basis offurther investigation using multiple imaging modalities.

Conclusion

The observation of an epiphyseal scar is a characteristic thathas traditionally been associated with individuals of a youn-ger chronological age due to the expected obliteration of theradio-opaque line by osseous remodelling. The results ofthis study suggest that although a statistically significantrelationship does exist between chronological age and per-sistence of the epiphyseal scar at the proximal and distaltibia, the extent of its influence may have been over-estimated in the literature pertaining to forensic age estima-tion. This study suggests that less than 5 % of variation inTPS can be attributed to variation in chronological age andis, therefore, insufficient to infer a causal relationship. It is,therefore, suggested that the observation of an epiphysealscar in a skeletally mature individual at either the proximalor distal tibia should not be used to infer a temporal prox-imity to epiphyseal fusion. Consequently, the presence of anepiphyseal scar should not be considered as a maturitycriterion in the estimation of chronological age.

References

1. Khan KM, Miller BS, Hoggard E, Somani A, Sarafoglou K (2009)Application of ultrasound for bone age estimation in clinical prac-tice. J Pediatr 154(2):243–247

2. Greulich WW, Pyle SI (1959) Radiographic atlas of skeletal de-velopment of hand and wrist. Stanford University Press, Palo Alto

3. Pyle SI, Hoerr NL (1969) A radiographic standard of reference forthe growing knee. Charles C. Thomas, Springfield

4. Tanner JM, Healy MJR, Goldstein H, Cameron N (2001)Assessment of skeletal maturity and prediction of adult height(TW3 method). Saunders, London

5. Mentzel H-J, Vilser C, Eulenstein M, Schwartz T, Vogt S, Böttcher J,Yaniv I, Tsoref L, Kauf E, Kaiser W (2005) Assessment of skeletalage at the wrist in children with a new ultrasound device. PediatrRadiol 35(4):429–433

6. Dedouit F, Bindel S, Gainza D, Blanc A, Joffre F, Rougé D,Telmon N (2008) Application of the Iscan Method to two- andthree-dimensional imaging of the sternal end of the right fourth rib.J Forensic Sci 53(2):288–295

7. İşcan MY, Loth SR, Wright RK (1984) Metamorphosis at thesternal rib end: a new method to estimate age at death in whitemales. Am J Phys Anthropol 65(2):147–156

8. Acsadi G, Nemerski J (1970) History of human life span andmortality. Akademiai Kiado, Budapest

9. Brooks S, Suchey J (1990) Skeletal age determination based on theos pubis: a comparison of the Acsádi–Nemeskeri and Suchey–Brooks methods. Hum Evol 5:227–238

10. Ritz-Timme S, Cattaneo C, Collins MJ, Waite ER, Schütz HW,Kaatsch HJ, Borrman HIM (2000) Age estimation: the state of theart in relation to the specific demands of forensic practise. Int JLegal Med 113(3):129–136

11. Koc A, Karaoglanoglu M, Erdogan M, Kosecik M, Cesur Y (2001)Assessment of bone ages: is the Greulich–Pyle method sufficientfor Turkish boys? Pediatr Int 43(6):662–665

12. Bull RK, Edwards PD, Kemp PM, Fry S, Hughes IA (1999) Boneage assessment: a large scale comparison of the Greulich and Pyle,and Tanner and Whitehouse (TW2) methods. Arch Dis Child 81(2):172–173

13. Cole AJL, Webb L, Cole TJ (1988) Bone age estimation: a com-parison of methods. Br J Radiol 61(728):683–686

14. O' Connor JE, Bogue C, Spence LD, Last J (2008) A methodto establish the relationship between chronological age andstage of union from radiographic assessment of epiphysealfusion at the knee: an Irish population study. J Anat 212(2):198–209

15. Baumann U, Schulz R, Reisinger W, Heinecke A, Schmeling A,Schmidt S (2009) Reference study on the time frame for ossifica-tion of the distal radius and ulnar epiphyses on the hand radio-graph. Forensic Sci Int 191(1–3):15–18

16. Groell R, Lindbichler F, Riepl T, Gherra L, Roposch A, Fotter R(1999) The reliability of bone age determination in centralEuropean children using the Greulich and Pyle method. Br JRadiol 72(857):461–464

17. Ding M, Odgaard A, Linde F, Hvid I (2002) Age-related variationsin the microstructure of human tibial cancellous bone. J OrthopRes 20(3):615–621

18. McCalden RW, McGeough JA, Court-Brown CM (1997) Age-related changes in the compressive strength of cancellous bone.The relative importance of changes in density and trabeculararchitecture. J Bone Joint Surg Br 79(3):421–427

19. Cope Z (1920) Fusion-lines of bones. J Anat 55:36–3720. Schaefer MC, Black SM (2005) Comparison of ages of epiphyseal

union in North American and Bosnian skeletal material. J ForensicSci 50(4):777–784

Int J Legal Med

21. Hoerr NL, Pyle SI, Francis CC (1962) Radiographic atlas ofskeletal development of the foot and ankle: a standard of reference.Charles C Thomas, Springfield

22. Crowder CM, Austin D (2005) Age ranges of epiphyseal fusion inthe distal tibia and fibula of contemporary males and females. JForensic Sci 50(5):1001–1007

23. Todd TW (1930) The anatomical features of epiphyseal union.Child Dev 1(3):186–194

24. Todd TW (1930) Comparative youth: the physical aspect. ChildDev 1(2):79–89

25. Cameriere R, Cingolani M, Giuliodori A, De Luca S, Ferrante L (2012)Radiographic analysis of epiphyseal fusion at knee joint to assess likeli-hood of having attained 18 years of age. Int J Legal Med:1–11

26. Kausar A, Varghese PS (2012) Estimation of age by epiphysealunion of the knee joint by radiological examination in Bijapurdistrict. Int J Biomed Adv Res 3(2):132–138

27. Weiss E, DeSilva J, Zipfel B (2012) Brief communication: radio-graphic study of metatarsal one basal epiphyseal fusion: a note ofcaution on age determination. Am J Phys Anthropol 147(3):489–492

28. Ostrum RF (2001) Techniques of retrograde intramedullary nailingof the femur. Tech Orthop 16(4):354–360

29. Ostrum RF, Storm S, White K (2009) The epiphyseal scar as aradiographic landmark for retrograde femoral nail insertion. Am JOrthop 38(9):442–444

30. Clement H, Heidari N, Kosuge D, Grechenig W, Tesch N,Weinberg A (2011) Anatomical structures at risk with the proudretrograde femoral nail. Arch Orthop Trauma Surg:1–6

31. Brumback RJ (1996) The rationales of interlocking nailing of thefemur, tibia, and humerus: an overview. Clin Orthop 324:292–320

32. The Scottish Government (2010) Relative poverty across Scottish localauthorities. The Scottish Government. http://www.scotland.gov.uk/Resource/Doc/322580/0103786.pdf. Accessed 7/2/12 2012

33. Walker RA, Lovejoy CO (1985) Radiographic changes in theclavicle and proximal femur and their use in the determination ofskeletal age at death. Am J Phys Anthropol 68(1):67–78

34. Kjar I (1974) Skeletal maturation of the human fetus assessedradiographically on the basis of ossification sequences in the handand foot. Am J Phys Anthropol 40(2):257–275

35. Ameen S (2005) Harris lines of the tibia across centuries: a com-parison of two populations, medieval and contemporary in CentralEurope. Skeletal Radiol 34(5):279–284

36. Velasco-Vázquez J, Martín-Rodríguez E, Arnay-de-la-Rosa M,González-Reimers E, Castilla-García A (1999) Harris lines in the

prehispanic population of Gran Canaria (Canary Islands). HumEvol 14(3):169–173

37. Cross N, Hillman L, Allen S, Krause G, Vieira N (1995) Calciumhomeostasis and bone metabolism during pregnancy, lactation, andpostweaning: a longitudinal study. Am J Clin Nutr 61(3):514–523

38. Porac C, Coren S, Duncan P (1980) Life-span age trends inlaterality. J Gerontol 35(5):715–721

39. Ozener B (2012) Extreme behavioral lateralization and the remod-eling of the distal humerus. Amer J Hum Biol:436–440

40. Blackburn A (2011) Bilateral asymmetry of the humerus duringgrowth and development. Am J Phys Anthropol 145(4):639–646

41. Auerbach BM, Ruff CB (2006) Limb bone bilateral asymmetry:variability and commonality among modern humans. J Hum Evol50(2):203–218

42. Ruff CB (1981) Bilateral asymmetry in cortical bone of the hu-merus and tibia—sex and age factors. Hum Biol 53(1):69

43. Molnar P (2006) Tracing prehistoric activities: musculoskeletalstress marker analysis of a stone-age population on the Island ofGotland in the Baltic sea. Am J Phys Anthropol 129(1):12–23

44. Hughes C, Heylings DJA, Power C (1996) Transverse (Harris)lines in irish archaeological remains. Am J Phys Anthropol 101(1):115–131

45. Langley-Shirley N, Jantz RL (2010) A Bayesian approach to ageestimation in modern Americans from the clavicle. J Forensic Sci55(3):571–583

46. Kreitner KF, Schweden FJ, Riepert T, Nafe B, Thelen M (1998)Bone age determination based on the study of the medial extremityof the clavicle. Eur Radiol 8(7):1116–1122

47. Schmeling A, Grundmann C, Fuhrmann A, Kaatsch HJ, Knell B,Ramsthaler F, Reisinger W, Riepert T, Ritz-Timme S, Rösing F,Rötzscher K, Geserick G (2008) Criteria for age estimation inliving individuals. Int J Legal Med 122(6):457–460

48. Schulz R, Mühler M, Mutze S, Schmidt S, Reisinger W, SchmelingA (2005) Studies on the time frame for ossification of the medialepiphysis of the clavicle as revealed by CT scans. Int J Legal Med119(3):142–145

49. Mühler M, Schulz R, Schmidt S, Schmeling A, Reisinger W (2006)The influence of slice thickness on assessment of clavicle ossificationin forensic age diagnostics. Int J Legal Med 120(1):15–17

50. Kellinghaus M, Schulz R, Vieth V, Schmidt S, Schmeling A (2010)Forensic age estimation in living subjects based on the ossification statusof the medial clavicular epiphysis as revealed by thin-slice multidetectorcomputed tomography. Int J Legal Med 124(2):149–154

Int J Legal Med