Dental cementum apposition as an indicator of age .TORE SOLHEIM Section for Forensic Odontology, Department of Pathology^ Dental Faculty, University of Oslo, Oslo, Norway
Solheim T: Dentai cementum apposition as an indicator of age. Scand J Dent Res 1990; 98: 510-9. Abstract — The purposes ofthe present investigation were to examine various ways of measuring cementum thickness, to study its relationship wdth age, and to consider what contribution this factor might offer in methods for age estimation. The amount of dental cementum apposition was studied in 1000 teeth, excluding molars, from a Caucasian population. Cementum thickness was estimated according to the scoring methods suggested by GUSTAFSON and by JOHANSON. In addition, the width of the cementum was measured at the apex and also at approximately one third of the root length from the apex. Statistical analyses using the SPSS package in a Cyber computer indicated a symmetric left/right distribution of cementum thickness. The sum of tbe cementum thickness on vestibular and lingual surfaces, measured at one third of the root length from the apex, showed the strongest correlation with age (r = 0.40 to 0.65). A reduced rate of cementum apposition was observed in the elderiy. Also, maxillary teeth had more cementum on the lingual than on the vestibular surfaces. A tendency was noted for less cementum to occur in women than in men and on teeth removed from deceased persons or extracted for pathologic reasons. Correlation coefficients indicated that, for at least some types of teeth, the cementum thickness might give a significant contribution to statistical methods ofage assessment. Key words: age determination by teeth; age factors; aging; dental cementum; sex charaeteristies. Box 1052 Blindem, N-0316 Oslo 3, Norway. Accepted for publication 20 April 1990.
Apposition of cementum on the surface of dental roots was introduced by GDSTAFSON (1) as one of six factors in his method for age estimation in forensie odontology. DALITZ (2), on the other hand, found that the amount of eementutn was not sufficiently related to age and discarded this factor in his method. Later JOHANSON (3) again used the atnount of cementum for age estimation.
Cementum is formed as a result of a eontinuous process throughout life and has been fotind to be tripled in thickness between the age of 20 and 60 yr (4). The thiekness varies, with a maximum at the apex and a minimum near the cementoenamel junction. While acellular in the coronal portion of the root, cementum in the apical half is both aeeliular and cellular. The latter form contains living
DENTAL CEMENTUM AS AN INDIGATOR OF AGE cells, cementocytes (5) and is typically found in hypercementosis. The continuous apposition of cementum may be influenced by function, occlusion and pathologic processes. Teeth affected by periodontal disease have been shown to have a thinner cementum than healthy teeth (6). Further, functional teeth are furnished with less cementum than non-occluding teeth (7), while impacted teeth seem to have less cementum than functional teeth (8). The aims of this investigation were to examine various ways in which the cementum thickness could be measured, to study the relationship between thickness and age, and to consider what contribution this factor might offer to methods for age estimation. It was also the intention to study the relationship between the widths of the cementtim zone and the sex, reason for extraction, periodontal destruction, translucent zone and root dentin color. Furthermore, multiple regression analysis for each type of tooth would be applied to find formulas useable for age prediction based on cementum thickness as the only factor.
511
the vesdbular (GE2) side at about one-third of the root length from the apex and then again on both sides at the apex (GE3, CE4) (Fig. 1). Tooth color (GEST) was estimated and translucericy (TSD) measured, as described elsewhere (9, 13). STATISTICAL ANALYSES The data were entered into a Gyber (CDG) computer for stadstical analyses, applying the Cyber version of the SPSS Package release 8.3 (12). A paired l-test for left/right differences was performed on contralateral teeth from the same individual (20-28 pairs of each tooth type). Since no differences were detected at P< 0'.05 level, contralateral teeth were pooled for the final analyses. As these operated with the tooth, and not the
Materials and methods A detailed description of the material, consisting of 1000 teeth, 100 of each type excluding molars, from a Caucasian population, has been given in a previous communication (9). A type of tooth is defined as comprising the two contralateral teeth of the same sort, e.g. maxillary second premolars. A tooth-age was calculated for each tooth by using the age of the individual and subtracting the mean age at which the root is completed (10). The surface of half-sectioned teeth (II) was stained with carbol fuchsin (1% basic fuchsin solution) for 3 min, rinsed, and polished with a finegrained sandpaper. The cementum was studied under a stereomicroscope and its aroount estimated according to the scores of GUST AFSON (1) (GEG) and of JoHANSON (3) (CEJ); Johanson's scores being doubled to avoid half-units. The thickness of the cementum was then measured in the stereomicroscope with a measuring eyepiece. Measurements were made on both the lingual (GEl) and
O i l 1 i\
f/
// l_^ Jb
j-'E' 1 X^l_3' r ^ r '-' "'' ^''^ Fig. I. Drawing indicating various points of measuring cementum thickness. GEl is measured at one-third of root length from apex and perpendkularly to dentitio-cemental junction on lingual side of tooth. GE2 is same measurement on vestibular side. CE3 is thickness at apex perpendicular to dentino-cemental juncdon on lingual side. GE4 is same measurement on vestibnlar side.
312
: SOLHEIM
indiviiduai, as the unit, only ohe from each pair of contralateral teeth from ajsingle individual could be included in .the study. The right, Eooth was chosen, which'reduced the number of teeth' for the fmal analyses from: :IOO to 80i'*:72 for the various t)'pes of teeth. The following transformations and ratios were computed: C]=CE1+GE2. CF1 = GE3/CE1, CF3 = C4/CI, CF5 = GI/CEJ, CEJ2 = GEJ**2, G42 = G4**2, LCEJ = LG10(CEJ),
G4=GE3 + GE4, GF2 = CE4/GE2, GF4 = C1/CEG, = GEe**2, = G1**2, = LG10(GEG), = LG1O(C1),
For each t)=pe of tooth a condescriptive analysis was ran. Squared and logarithmic (base 10 logarithm) transformations of the data were calculated by the computer, in order' to test for a non-linear
relationship between age and different measurements of cementum. The relationships between different methods of measuring the cementum and age, tooth age, the sex and reason for extraction were analyzed, using the Pearson and Speartnan rank correlation coefEcients. Partial correlation analyses were run for the relationship between age and cementum thickness, controlling for the effects of the sex and reason for extraction. Also, partial correlation analyses were run for the relationship 'between cementum and the sex and the reason for extraction, controlling for the effect of age. Finally, partial correlation analyses were run for the relationship between cementum and estimate of tooth color (GEST) and measured translucency (TSD). Multiple regression analyses were also run, using age as the dependent variable. A stepwise procedure with inclusion level P< 0.05 was chosen and a formula for age prediction calculated, In addition, a run was made with tooth age used as
Table 1 Mean values of cementum thickness
GEG GEJ GEl GE2 GE3 GE4 Ma:x. 1 Max. 2 Max. 3
Max. 4 Max. 5 Mand. Mand. Mand. Mand.
I 2 3 4
Matid. 5
1.2 1.2 1.6 1.4 1.4
2.7 2.7 3.4 3.1 3.1
189 114 21,7
1.2 1.2 1.4 1.4 1.4
2.8 2.5 3.1 3.1 3.1
102 126 124 113 141
164 166
GI
G4
GFl
GF2 GF3 GF4 GF5
161 116 167 124 153
388 387 650 433 415
258 292 365 348 365
349 230 384 289 320
618 663 995 780 780
2.1 3.5 3.5 3.1 2.9
1.6 2,7 2.2 3.2 2.6
1.8 3.0 2.8 3.0 2.7
111
286 264 529 546 521
303 254 480 549 399
214 569 250 470 258 1002 218 1087 270 920
2.7 2.0 4.5 5,2 4.0
2.6 2.0 3.7 5.8 3.2
2.6 2.0 4.0 5.4 3.5
124 138 107 129
CEG = cementum according to GDSTAFSON (1) CEJ = cementum according to JOHANSON (3). GEl = thickness of cementum in jim at 1/3 of root length from apex on lingual side. GE2= thickness of cementum in |lm at 1/3 of rooth length from apex on vestibular side. GE3 = thickness of cementum in |im apically on lingual side. GE4 = thickness of cementum in (im apically on vestibular side. GI =GEi+GE2. G4 =GE3 + GE4. GFl =GE3/GE1. GF2 =GE4/GE2. GF3 =G4/G1. CM =G1/GEG. 'GF5 =G1/GEJ.
302 197 254 200 237
133 89 114 92 105
183
83 108 87 76 92
200 190 163 196
DENTAL.CEMENTUM AS AN INDICATOR OF AGE the dependent variable. This regression was also forced through origin (age = O) in an extra ron.
Results The mean amount of cementum aceording to difierent methods of measurement is shown in Table 1. The cementum was from two to three times thieker at the apex than at one-third ofthe root length from the apex. Mandibular canines revealed more cementum than did any of the other teeth. This was particularly noticeable on the lingual side. Maxillary canines, however, had about the same amount of cementum as maxillary premolars. Lingual surfaees tended to be covered with more cementum than were vestibular surfaces in maxillary teeth. In tnandibular teeth, sueh a tendency was found in canines and second premolars only. Pearson correlation coefficients (Table 2) between age and amount of cementum showed that the sum of lingual and vesdbular thicknesses measured at one-third of the root
513
length from apex (Cl) displayed the strongest correlation (0.31-0.72). For all teeth except maxillary premolars this correlation was stronger than that between age and the scores of JOHANSON (CEJ) (0.22-0.48). For
all teeth except maxillary second premolars the corresponding correlations for cementum thickness at the apex (CE3, CE4r, and d) were weaker than for Cl. Also, the indices of CFl to CF5 had weaker correlation than Cl. Logarithmic transformation of Cl, however,, showed a stronger correlation with age than the untransformed Cl did for some teeth. Spearm.an rank correlation for Cl was, except for maxillary premolars, near the values for Pearson correlation. In Fig. 2 the increasing thickness of cementum at the level of one-third of root length from the apex is shown for three teeth of different ages. Scattergram of the CEl distribution versus age (Fig. 3) indicated a near-normal distribution. Partial correlation between age and amount of eementum, controlling for the sex
Table 2 Pearson correlation between age aad cementum thickness CEG
CEJ
CEl
CE2
CE3
CE4
Cl
C4
C12
C42
LCI
LC4 RSCl RSC4
Max.l Max. 2 Max. 3 Max. 4 Max. 5
0.31 0.07 0.40 0.38 0.26
0.38 0.22 0.45 0.46 0.40
0.4! 0.68 0.48 0.36 0.24
0.40 0.65 0.60 0.43 0.34
0.16 0.25 0.19 0.40 0.34
0.28 0.29 0.27 0.20 0.20
0.49 0.72 0.58 0.44 0.31
0.16 0.25 0.29 0.36 0.31
0.47 0.70 0.51 0.35 0.20
0.15 0.23 0.26 0.30 0.23
0.49 0.70 0.61 0.50 0.42
0.15 0.22 0.27 0.38 0.38
0.52 0.74 0.61 0.48 0.45
0.22 0.27 0.32 0.41 0.41
Mand. 1 Mand. 2 Mand. 3 Mand. 4 Mand. 5
0.35 0.39 0.38 0.24 0.37
0.48 0.40 0.39 0.30 0.41
0.63 0.48 0.53 0.59 0.47
0.58 .0.55 0.57 0.61 0.53
0.38 0.36 0.66 0.32 0.30 0.44 0.54- 0.16 0.32 0.46 0.58 0.41 0.20 0.16 0.59 0.15 0.31 0.51 0.52 0.46
0.60 0.46 0.50 0.49 0.34
0^27 0.18 0.41 0.14 0.40
0.69 0.6! 0.63 0.64 0.63
0.11 0.06 0.40 0.18 0.49
0.65 0.56 0.62 0.67 0.56
0.30 0.15 0.3B 0.16 0.49
Abbreviations: see Table 1. G12 = square of Gi C42 = square of C4 LCI = logarithmic transformation of Gl (base 10 logaritlim) LG4 = logarithmic transformation of G4 (base 10 logarithm) RSGl = Spearman correlation for Cl and age RSG4 = Spearman correlation for 0 4forage Significapce limits: P0.22 (H = 78) f 0.29 ()i=78)
514
SOLHEIM
(Table 3), showed almost identical coefficients as compared to the simple Pearson correlation. Partial correlation coefficients between amount of cementum (Cl) and the sex, controlling for age, were significantly positive for second premolars -and maxillary canines alone. A positive but not significant correlation was, however, also found for all the other t)rpes of teeth. Multiple regression analysis (Table 4) resulted in quite different formulas-for the different types of teeth. The strongest correlation for the formulas varied from r = 0.54 for maxillary incisors to r = 0.75 for mandihular second premolars. For some teeth the toothage showed stronger correlation than did the individual's age; the differences were, however, small and insignificant. For most teeth the Cl made the strongest contribution to
the regression. For all teeth except maxillary laterals and second premolars, the square of Gl (C12) added significantly to the regression, indicating a reduced speed of cementum apposition in older persons. For all teeth except maxillary laterals and premolars, the sex was added to the regression with a negative sign. Since men were indicated by 1 and women by 0, this indicated more cementum in men. For some teeth the factors EXl, EX2, EX3 significantly contributed to the regression with a positive sign. These findings indicted a reduced amount of cementum in teeth extracted for pathologic reasons or from corpses. A comparatively strong correlation was found between the amount of cementum (Cl) and the extent of the translucent zone (TSD) and also between the former and the
Fig. 2. Stereomicroscopic photographs of cementum thickness at lingual surface one-third from ape;(GEl) in teeth from patients aged 27 yr (A), 48 yr (B), and 75 yr (C). C, cementuiu and D, deniiu Cementum thickness is shown between arrows. Bar = 200 |im.
DENTAL GEMENTUM AS AN INDIGATOR OF AGE
515
,AGE VERSUS CEMENTUM THICKNESS
o o-
S-
X X
o LU
X
X X
X X
X X
LU
(3