Liversidge 2010.3


16

Many studies of dental maturation during the last 
50 years have described the timing of permanent tooth 
formation stages. Several reports remain important 
because they include very young children and follow 
individuals longitudinally (Moorrees et al., 1963) or use 
clearly defined stages and a large sample (Demirjian et 
al., 1973; Demirjian and Goldstein, 1976; Demirjian and 
Levesque, 1980; Demirjian, 1994), although none give full 
descriptive results. The ease of statistical analyses and a 
better understanding of age estimation have highlighted 
the lack of descriptive data of the timing of tooth 
formation stages. This paper presents detailed results 
from a collaboration of published cross-sectional studies 
organised by Nils Chaillet in Canada that has resulted in 
several published reports including a polynomial approach 
to Demirjian’s dental maturity scale (Chaillet et al., 2004; 
Chaillet et al., 2005) and maturity data of individual tooth 
stages (Liversidge et al., 2006).  Tooth formation data are 
presented here in different formats to assess maturity and 
estimate age using developing teeth in living children, 
in forensic cases, or in archaeological cases where sex is 
uncertain.

MATERIAL AND METHODS

Tooth formation data from dental radiographs using 
Demirjian stages (Demirjian et al., 1973; 1976; Demirjian, 
1994) were combined to form the International Data Base. 
The sample consisted of cross-sectional data from published 
studies from Finland, Sweden, England, Korea, Belgium, 
Australia, Belgium, Canada and France (Nyström et al., 
1986; Nyström et al., 1988; Kataja et al., 1989; Liversidge 
and Speechly, 2001; Teivens and Mörnstad, 2001; Willems 
et al., 2001; McKenna et al., 2002; Chaillet and Demirjian, 

Demirjian Stage Tooth Formation Results from a Large 
Group of Children
Helen M. Liversidge

Queen Mary University of London, Barts and The London School of Medicine and Dentistry, London, United 
Kingdom

Correspondence to:  Helen M. Liversidge, Institute 
of Dentistry, Turner Street, E1 2AD, London, United 
Kingdom
Email: h.m.liversidge@qmul.ac.uk

ABSTRACT    The aim of this study is to present further 
data on the timing and variation of individual permanent 
mandibular teeth using Demirjian stages from a large 
collaboration. Seven mandibular permanent teeth were 
assessed from dental radiographs of healthy dental 
patients from Australia, Belgium, Canada, England, 
Finland, France, South Korea and Sweden (cross-sectional 
study; n = 9,371, 4,710 males, 4,661 females; aged 2–18).  
Data are presented in three ways, namely by tooth stage 
for males, females, and pooled sex. Mean age at entry of 
each tooth formation stage (maturity data) was calculated 

using logistic regression and modified for age prediction. 
The 51% confidence interval for age within stage of 
individual tooth stages was calculated for use in forensic 
age estimation where the burden of proof is on the 
balance of probabilities. Average age, standard deviation, 
standard error, 3rd and 97th percentile within tooth stage 
was calculated from a uniform age sample (171 for each 
year of age from 3 to 16, n = 2,394). Modified maturity 
data and average age within stage from the uniform age 
distribution are two new methods of age estimation. 
Dental Anthropology 2010;23(1):16-23.

2004).  The sample studied in this paper, after cleaning was 
radiographic data from 4,710 males and 4,661 females aged 
2 to 18 (when all individuals had reached second molar 
maturity) shown in Figure 1 (left).  Previous results of the 
timing of Demirjian tooth stages of individual teeth are 
available for males and females for each group separately 
(Finland, Sweden, England, Korea, Belgium, Australia, 
Belgium, Canada and France) and for all groups combined 
for individuals from age 2 up to and including 16 years of 
age (Table 9 in Liversidge et al., 2006).

Mean age of entering a tooth stage was calculated by 
logistic regression for males, females and sexes combined 
(Table 1). Logistic regression calculates the average age 
at entry of a specific formation stage and represents the 
age when half of children at that age, have reached or 
passed the stage (Taranger, 1976; Eveleth and Tanner, 1990; 
Cameron, 2004) and is similar to probit regression used to 
calculate mean age of tooth eruption (Liversidge, 2003).  In 
this regard the mean age is identical to median age, half of 
children enter the stage prior to mean age, and half enter 
subsequent to mean age. This is an appropriate method to 
compare maturation between groups; it is not equivalent to 
the mean or median age of a child in the specific maturity 
stage (see below). Maturity data were modified for each 
stage of each developing tooth (Table 2) by adding half the 
interval to the next stage (see Smith, 1991).  The second way 
of presenting results is the 51% confidence interval of age 



17

for each individual tooth stage. This was calculated using 
the 24.5th and 75.5th percentiles of average age within 
stage (Table 10 from Liversidge et al., 2006) and is shown in 
Table 3. This is slightly greater than the inter-quartile range 
and just over half of individuals in the specific stage will 
fall within this interval. The third way data are presented is 
average age within stage from a uniform age sample.  The 
age cohort with the lowest number of children was 171 for 
three year olds (87 male, 84 female) and 87 males and 84 
females were randomly selected from each year of age from 
3 to 16 (total n 2,394, Figure 1 right). Descriptive statistics 
of age within individual tooth formation stages, including 
3rd and 97th percentiles (using the normal deviate, see 
Cole, 2002) were calculated from this group and are shown 
in Table 4. This is referred to as L10a in Liversidge et al. 
(2010) in a comparison of dental age estimation methods 
using the same target sample of Maber et al. (2006). 

RESULTS AND DISCUSSION

Maturity data representing the average age entering 
a tooth formation stage (age when half of children have 
reached or passed the stage) are shown in Table 1. The 
ages when 3% and 97% of girls and boys had entered some 
stages including D (crown complete with initial root) and 
H (mature apex) are shown in Figure 2. The left and right 
hand edges of the open diamond are the ages when 3% 
and 97% of girls had reached or passed this stage. Filled 
diamonds are data for boys. The sex difference is smaller 
for earlier stages and greatest for stage H (mature apex) 
of the canine.  This is the only appropriate measure of 
the final maturity stage. Smoothed cumulative frequency 
distribution curves are shown for M2 stages in Figure 3. 
This is the only tooth in this study where data are available 
from crypt stage to mature apex; however the variation 
for early tooth stages is probably unrepresentative as 

Fig. 1. Age and sex of radiographic sample (left), selected group for a uniform age distribution (right).

Fig. 2. Timing of some Demirjian tooth stages including 
D (crown complete with initial root) and H (mature apex). 
Diamond extends from 3rd to 97th percentile, apex is mean 
age. Open diamonds females, solid diamonds males.

this collaborative study includes only 38 two year olds. 
Maturity data modified for age prediction are shown in 
Table 2 (referred to as L9a in Liversidge et al., submitted). 
Once a tooth reaches the most mature stage (apex closed 
with mature periodontal ligament width) age cannot be 
estimated using development and this stage is omitted 
from tables of modified maturity data and within stage 
data. The second type of result is the 51% confidence 
interval (Table 3, Figure 4) similar to what Koningsberg 
et al. (2008) term ‘coverage’. This interval is useful when 
estimating age in forensic cases where the burden of 
proof is on the balance of probabilities. Coverage also 

DEMIRJIAN TOOTH FORMATION STAGES



18

 Males Alone Females Alone Sexes Pooled
Mandibular Mean for  Mean for  Mean for
 Tooth Grade prediction sd prediction sd prediction sd

 I1 E 4.39 0.44 4.24 0.42 4.33 0.43
  F 6.02 0.38 5.70 0.36 5.88 0.38
  G 6.97 0.40 6.68 0.37 6.82 0.39
  H 8.35 0.60 7.84 0.57 8.09 0.60
 I2 E 5.21 0.47 4.87 0.42 5.05 0.46
  F 6.68 0.38 6.28 0.37 6.49 0.39
  G 7.82 0.47 7.32 0.44 7.56 0.47
  H 9.41 0.65 8.79 0.59 9.09 0.65
 C D 4.75 0.59 4.19 0.53 4.49 0.59
  E 6.60 0.45 5.99 0.44 6.31 0.47
  F 8.51 0.51 7.59 0.50 8.04 0.56
  G 11.04 0.61 9.84 0.56 10.42 0.67
  H 13.23 0.64 11.73 0.68 12.47 0.80
 P1 C 3.58 0.40 3.41 0.35 3.49 0.39
  D 5.56 0.43 5.28 0.41 5.43 0.43
  E 7.26 0.49 6.78 0.43 7.02 0.48
  F 9.19 0.53 8.70 0.49 8.94 0.53
  G 11.44 0.62 10.70 0.58 11.05 0.63
  H 13.03 0.64 12.27 0.65 12.64 0.69
 P2 B 3.81 0.57 3.80 0.55 3.81 0.56
  C 4.94 0.55 4.81 0.55 4.89 0.55
  D 6.48 0.57 6.21 0.56 6.35 0.57
  E 8.02 0.58 7.62 0.64 7.82 0.62
  F 9.83 0.69 9.35 0.60 9.58 0.65
  G 12.26 0.75 11.57 0.71 11.90 0.76
  H 14.09 0.73 13.44 0.76 13.77 0.78
 M1 E 3.99 0.32 3.81 0.38 3.91 0.35
  F 5.46 0.41 5.21 0.41 5.35 0.42
  G 6.84 0.48 6.42 0.45 6.64 0.46
  H 9.95 0.71 9.33 0.68 9.64 0.71
 M2 O 3.40 0.30 2.92 0.51 3.19 0.40
  A 3.46 0.33 3.33 0.42 3.40 0.37
  B 3.97 0.33 3.72 0.46 3.87 0.38
  C 4.91 0.44 4.85 0.44 4.88 0.42
  D 6.79 0.52 6.43 0.54 6.61 0.54
  E 8.80 0.56 8.48 0.52 8.64 0.55
  F 10.68 0.60 10.12 0.56 10.39 0.60
  G 12.18 0.60 11.57 0.60 11.86 0.63
  H 15.22 0.71 14.75 0.77 14.99 0.75

TABLE 1. Mean age of entering Demirjian tooth stages for mandibular permanent teeth recalculated from cleaned data using 
logistic regression of 4,710 males and 4,661 females1

 1These statistics represent the age when half of children have reached or passed the stage. Code O refers to crypt 
formation prior to evidence of tooth mineralization.

provides a means to test if a target sample is representative 
(Konigsberg et al., 2008); 24.5% should fall below the 
interval, 51% within the interval and 24.5% above. The 
third type of result is average age within stage from a 
uniform age distribution (similar number of individuals 
in each age category). Konigsberg and Frankenberg (1992) 
suggest that age could be more accurately estimated using 
a reference sample based on a uniform age distribution. 

Results of this type from this study are shown in Table 4 
and are referred to as L10a in Liversidge et al. (submitted). 
This flat age distribution, together with an adequate 
sample size and wide age range, are important features of 
a reference sample that contribute to accuracy (Konigsberg 
and Frankenberg, 2002; Konigsberg et al., 2008). 

Understanding how maturity data differ to within stage 
data is challenging and Figure 5 illustrates some of these 

H.M. LIVERSIDGE.



19

 Males Alone Females Alone Sexes Pooled
Mandibular Mean for  Mean for  Mean for
 Tooth Grade prediction sd prediction sd prediction sd

 I1 E 5.24 0.44 5.01 0.42 5.11 0.43
  F 6.48 0.38 6.18 0.36 6.35 0.38
  G 7.64 0.40 7.29 0.37 7.45 0.39
 I2 E 5.94 0.47 5.62 0.42 5.77 0.46
  F 7.20 0.38 6.86 0.37 7.02 0.39
  G 8.56 0.47 8.06 0.44 8.32 0.47
 C D 5.68 0.59 5.09 0.53 5.40 0.59
  E 7.64 0.45 6.83 0.44 7.18 0.47
  F 9.86 0.51 8.76 0.50 9.23 0.56
  G 12.14 0.61 10.80 0.56 11.44 0.67
 P1 C 4.61 0.40 4.37 0.34 4.46 0.39
  D 6.34 0.43 5.99 0.41 6.22 0.43
  E 8.20 0.49 7.82 0.43 7.98 0.48
  F 10.32 0.53 9.83 0.49 10.00 0.53
  G 12.24 0.62 11.49 0.58 11.84 0.63
 P2 B 4.41 0.57 4.30 0.54 4.34 0.56
  C 5.62 0.55 5.46 0.55 5.62 0.55
  D 7.16 0.57 6.86 0.56 7.08 0.57
  E 8.90 0.58 8.48 0.64 8.70 0.62
  F 11.04 0.69 10.50 0.60 10.74 0.65
  G 13.19 0.75 12.56 0.71 12.84 0.76
 M1 E 4.80 0.32 4.58 0.38 4.63 0.35
  F 6.20 0.41 5.81 0.41 6.00 0.42
  G 8.38 0.48 7.84 0.45 8.14 0.46
 M2 O 3.43 0.30 3.12 0.51 3.30 0.40
  A 3.65 0.33 3.74 0.42 3.64 0.37
  B 4.35 0.33 4.74 0.46 4.38 0.38
  C 5.59 0.44 5.84 0.44 5.75 0.42
  D 7.45 0.52 7.78 0.54 7.62 1.05
  E 9.35 0.56 9.74 0.52 9.52 1.10
  F 10.84 0.60 11.46 0.56 11.13 1.15
  G 12.92 0.60 13.74 0.60 13.42 1.20

TABLE 2.  Maturity data modified for age prediction for mandibular permanent tooth stages

differences.  Smoothed cumulative distribution curves for 
stages D, E and F of M2 (sexes pooled) are shown. These 
curves represent the increasing proportion of children at 
each age who have reached or passed the specific stage.  
A tooth is considered to be ‘in’ a stage until it enters the 
next stage. The shaded area shows the age interval of all 
individuals within stage D; ranging from the youngest 
(most dentally advanced) to the oldest individual in that 
stage (most dentally delayed). Maturity is a continuum and 
we arbitrarily divide this into discrete stages, even though 
the process of maturation is gradual.  Stage D in molars is 
defined as crown complete with initial root spicules visible 
at the mesial and distal edges. These root spicules increase 
in length and the root bifurcation becomes visible, firstly 
as a dot or line, then as a semi-lunar radio-opacity. Once 
this occurs, the tooth is deemed to be in stage E. The three 
types of tooth data in this study are summarised in Figure 
5. Mean age at entry for M2 stage D is shown as a dot, 

maturity data modified for age prediction for this stage 
is shown as a triangle.  The age interval for individuals 
‘in’ stage D, extends from the youngest child in stage D, 
up to the age when the most delayed child leaves this 
stage and enters the next stage (when all individuals have 
entered stage E).  The age range of individuals within 
stage D and E are marked. The 51% age coverage for stage 
D is also shown. This figure was chosen for the forensic 
odontologist where the burden of proof is ‘on the balance 
of probabilities.’ This is an expression of the probability 
of estimated age being on one side of an age threshold.  
If an individual of unknown age presents with a second 
permanent molar in stage D (crown complete with initial 
root), the 51% age interval (from Table 3) is 7.01 to 8.50. On 
the balance of probabilities, the age of this individual is 
older than six but younger than nine. 

The existence of population differences in dental 
maturity is unclear and uncertain. Many studies report 

DEMIRJIAN TOOTH FORMATION STAGES



20 H.M. LIVERSIDGE.

 Males (n = 4,710)       Females (n = 4,661)      Combined (n = 9,371)
 Tooth Grade n 24.5% 75.5% n 24.5% 75.5% n 24.5% 75.5%

 I1 D 222 3.62 4.70 162 3.49 4.48 384 3.54 4.58
  E 425 4.73 5.91 297 4.58 5.60 722 4.67 5.80
  F 308 6.05 7.03 269 5.80 6.83 577 5.98 6.99
  G 561 7.25 8.57 460 7.01 8.09 1021 7.10 8.30
 I2 C 42 3.38 4.21 29 2.83 4.05 71 3.06 4.10
  D 386 4.08 5.20 253 3.80 4.95 639 3.98 5.03
  E 425 5.39 6.60 335 5.10 6.19 770 5.23 6.46
  F 428 6.96 8.00 353 6.50 7.64 781 6.72 7.95
  G 684 8.01 9.54 643 7.67 8.97 1327 7.90 9.10
 C C 327 3.85 5.01 173 3.52 4.62 500 3.71 4.93
  D 512 5.00 6.51 370 4.72 5.96 882 4.92 6.27
  E 745 6.99 8.40 542 6.46 7.78 1287 6.79 8.06
  F 1063 8.97 10.60 995 8.00 9.58 2058 8.43 10.06
  G 800 11.02 12.97 817 10.00 11.65 1617 10.44 12.30
 P1 A 20 2.57 3.34    21 2.59 3.33
  B 69 3.33 4.26 55 3.01 3.79 124 3.10 4.07
  C 433 4.38 5.44 308 4.18 5.28 741 4.28 5.40
  D 554 5.95 7.20 406 5.60 6.95 960 5.78 7.05
  E 803 7.67 9.00 795 7.20 8.53 1598 7.47 8.96
  F 918 9.40 11.00 889 8.90 10.54 1807 9.05 10.92
  G 580 11.32 12.97 657 10.63 12.10 1237 10.96 12.70
 P2 O 10 2.39 4.92    16 2.67 4.52
  A 89 3.54 4.72 77 3.30 4.44 166 3.45 4.60
  B 222 4.27 5.21 160 4.19 5.33 382 4.23 5.23
  C 446 5.17 6.68 338 5.06 6.48 784 5.10 6.58
  D 570 6.70 8.10 494 6.61 7.96 1064 6.66 8.01
  E 769 8.06 9.87 744 7.86 9.05 1513 7.99 9.46
  F 945 9.98 11.97 969 9.64 11.39 1914 9.83 11.76
  G 626 12.00 13.98 708 11.32 13.29 1334 11.70 13.74
 M1 D 131 3.31 4.22 100 3.22 4.03 231 3.25 4.14
  E 358 4.47 5.42 247 4.25 5.23 605 4.39 5.38
  F 432 5.66 9.53 313 5.42 6.56 745 5.56 6.90
  G 1279 7.70 9.53 1187 7.28 9.00 2466 7.50 9.20
 M2 O    19 2.89 3.91 28 3.11 4.28
  A 66 3.68 4.49 42 3.54 4.56 108 3.60 4.52
  B 215 4.31 5.05 175 4.08 5.00 390 4.20 5.02
  C 562 5.39 6.78 402 5.28 6.65 964 5.32 6.72
  D 799 7.10 8.70 791 6.98 8.30 1590 7.01 8.50
  E 794 8.98 10.48 734 8.57 10.01 1528 8.80 10.25
  F 572 10.55 12.03 629 10.04 11.59 1201 10.25 11.97
  G 983 12.53 14.50 1096 11.97 13.98 2079 12.02 14.11

TABLE 3.Fifty one percent coverage for mandibular tooth stages1

1This age interval includes 51% of individuals within each stage. Stage included if n ≥ 10

significant differences in average age within tooth stage 
between groups, but this is inappropriate to compare 
maturity or average age at entry. For example, imagine if 
the minimum age of a study was 7 and the average age at 
entry of M2 stage D is calculated. Looking at Figure 5, this 
will exclude many individuals younger than 7 who have 
entered this stage.  Looking at the cumulative incidence 
curve for stage D, it is clear that more than half of 7 years old 
in this large study have M2 in stage D (or later),  indicating 

that a minimum age of 7 is too old to calculate the average 
age for this stage.  Significant differences between groups 
have also been shown using a single dental maturity score 
such as Nolla (Nolla, 1960) or Demirjian (Demirjian et al., 
1973; 1976; Demirjian, 1994). These have been interpreted 
as due to either a secular trend or regional differences 
in dental maturation but little attempt has been made to 
investigate this more fully. There is little doubt that these 
differences occur, but it is questionable if they have any 



21

differences and similarities indicate, but no clear pattern or 
developmental shift is evident.  

Most studies of dental maturity are based on children 
of European origin. Mean age at entry for individual tooth 
stages was compared in the groups of this collaborative 
study. Children in Canada, Finland, Sweden, Belgium, 
England, France, Australia and a small group from 
Korea showed a wide age range of individuals within 
each tooth stage with many similarities in average age at 
entry (Liversidge et al., 2006). No single tooth type was 
consistently earlier or later and no clear pattern emerged 
from this analysis.  A comparison of mean age at entry 
of tooth stages by Moorrees et al. (1963) on White and 
Bangladeshi children aged 2 to 22 in London, United 
Kingdom showed few significant differences in mean age 
between these ethnic groups (Liversidge 2009).  The lack 
of published data of dental maturity from other regions 
of the world is sparse, but being addressed. Preliminary 
results from a worldwide comparative study showing 
stage H (apex mature) of the mandibular first molar in girls 
is shown in Figure 7. Smoothed cumulative distribution 
curves and 95% confidence interval of mean age (calculated 
from one year age groups using probit regression) are 
shown from the following regions: Australian Aboriginal 
(Liversidge and Townsend, 2006), Inuit, Japanese, Maori 
and Pacific Islanders (Moananui et al., 2008), South African 
Black and Cape Coloured, UK, Bangladeshi and White 
(Liversidge, 2009). The most advanced girls (youngest) 
girls in this stage are 6 years old and by 12, almost all have 
reached this stage. The average age at entry, when 50% of 
girls have reached this stage, is similar between groups, 
although two groups are slightly earlier than the others. 

These similarities between the reference study and the 
present study and between world groups are supported 
by recent histological findings in the duration of crown 
formation (Reid and Dean 2006) between the past and 
the present. For instance, molar crowns take around 3 
to 3.4 years to develop in maxillary, mandibular, first, 
second and third molars from Medieval Danes, northern 
European, South African Black and North American 
groups (Reid and Dean, 2006). The largest permanent 
crowns are found in Australian aborigines and the duration 
of enamel formation in molars in three first molars from 
this group is from 3 to 3.5 years (pers. comm. DJ Reid).  
Despite little documentation of the rate of dentine growth 
and root formation between individuals or groups, these 
findings suggest that the time it takes to grow a tooth is 
similar across time and between groups, especially at the 
resolution of crown and root fractions from radiographs. 

Assessing maturity or estimating age from crown 
and root stages is usually used for an individual child. In 
this regard the individual is compared to reference data 
while, in human biology, differences between groups are 
of interest. Small differences in the mean age of individual 
tooth stages at the group level  have little influence on 
the estimated age for an individual. Population specific 
reference data of radiographic tooth stages may be 

Fig. 3. Smoothed cumulative distribution curves 
(proportion of individuals and age) for stages crypt to H for 
the permanent mandibular second molar (sexes pooled).

Fig. 4.  Age interval of 51% coverage for stages A to G 
of permanent teeth (pooled sex).

biological meaning. 
The time interval between the original study Demirjian 

et al. (1973) and the present collaboration is between twenty 
to forty years.  Mean age at entry of individual tooth 
stages in girls from the Canadian reference (triangles apex 
at bottom; Demirjian and Levesque, 1980) and this study 
(triangles apex at top) are shown in Figure 6. The largest 
differences occur as a single stage difference in four teeth; 
the earliest stage with data for both incisors and canine 
and the second stage for first premolars with mean age in 
the recent study being later than the original reference. The 
mean ages in subsequent stages in these teeth are similar 
or close in age.  A comparison of the most mature stage in 
all tooth types is similar or marginally earlier in the recent 
study compared to the reference. It is unclear what these 

DEMIRJIAN TOOTH FORMATION STAGES



22

108-140.
Chaillet N, Demirjian A. 2004. Dental maturity in South 

France: A comparison between Demirjian’s method 
and polynomial functions. J Forensic Sci 49:1059–1066.

Chaillet N, Nyström M, Demirjian A. 2005. Comparison of 
dental maturity in children of different ethnic origins: 
International maturity curves for clinicians. J Forens Sci 
50:1164-1174. 

Cole TJ. 2002. Growth references and standards. In: 
Human growth and development. Cameron N, editor. 
San Diego: Academic Press. p 383-413.

Demirjian A. 1993-94. Dental development. CD-ROM, 
Silver Platter Education. Montreal: University of 
Montreal.

Demirjian A, Goldstein H. 1976. New systems for dental 
maturity based on seven and four teeth. Ann Hum Biol 
3:411-427.

Demirjian A, Goldstein H, Tanner JM. 1973. A new system 
of dental age assessment. Hum Biol 45:211-227. 

Demirjian A, Levesque GY. 1980. Sexual differences in 
dental development and prediction of emergence. J 
Dent Res 59:1110-1122.

Eveleth PB, Tanner JM. 1990. Worldwide variation in 
human growth. Cambridge: Cambridge University  
Press.

Kataja M, Nyström M, Aine L. 1989. Dental maturity 
standards in southern Finland. Proc Finn Dent Soc 
85:187-197.

Konigsberg LW, Frankenberg SR. 1992. Estimation of age 
structure in anthropological demography. Am J Phys 
Anthropol  89:235-256.

Konigsberg LW, Herrmann NP, Westcott DJ, Kimmerle EH.  
2008. Estimation and evidence in forensic anthropology: 
age-at-death. J Forensic Sci  53:541-557.

Fig. 5. A comparison of maturity data and within stage 
data. Smoothed cumulative distribution curves for stages 
D, E and F for second molars are shown.  Mean age of 
attainment and modified for age prediction are shown 
as dot and triangle respectively. Shaded area is the age 
interval of individuals within stage D. Age range within 
stage D and E are shown as well as 51% age coverage for  
stage D.

unnecessary and until evidence is available to show 
otherwise, the methods of age estimation presented here 
are appropriate for individuals from all groups.

CONCLUSIONS

New data on the timing of Demirjian stages from a 
large sample are presented in several ways. Two of these 
(modified maturity data, average age within stage from 
a uniform aged group) are new methods appropriate to 
estimate age from individual permanent teeth. The 51% 
coverage age interval for individual tooth stages are 
described for forensic age estimation where the burden 
of proof is on the balance of probabilities. These results 
represent the biggest data set and therefore probably 
the most reliable maturity reference and age estimation 
methods for mandibular permanent tooth formation using 
Demirjian stages in humans. 

ACKNOWLEDGMENTS

I am particularly grateful to Nils Chaillet (CHU Sainte-
Justine Research Centre, University of Montreal, Montreal, 
Quebec, Canada) for inviting collaboration and also 
Marjatta Nyström, Hakan Mörnstad, Kian Rowlings, Jane 
Taylor, Guy Willems and their co-authors for sharing data. 

LITERATURE CITED

Cameron N. 2004. Measuring maturity. Methods in Human 
growth research. In: Hauspie RC, Cameron N, Molinari 
L, editors. Cambridge: Cambridge University  Press. p 

Fig. 6. A comparison of mean age entering tooth stages 
for girls from Demirjian and Levesque (1980) triangles 
apex down and Liversidge et al. (2006) triangles apex up.

H.M. LIVERSIDGE.



23

 
M

al
es

 (
n

 =
 1

,2
18

) 
Fe

m
al

es
 (

n
 =

 1
,1

76
) 

Se
xe

s 
C

om
bi

n
ed

 (
n

 =
 2

,3
94

)
To

ot
h

 
St

ag
e 

n
 

m
ea

n
 

sd
 

3r
d

 
97

th
 

n
 

m
ea

n
 

sd
 

3r
d

 
97

th
 

n
 

m
ea

n
 

sd
 

se
m

 
3r

d
 

97
th

I1
 

D
 

11
1 

3.
96

 
0.

61
 

2.
81

 
5.

11
 

10
8 

3.
85

 
0.

57
 

2.
78

 
4.

92
 

21
9 

3.
90

 
0.

59
 

0.
04

 
2.

79
 

5.
01

 
E

 
14

4 
5.

18
 

0.
88

 
3.

53
 

6.
83

 
12

0 
5.

01
 

0.
81

 
3.

49
 

6.
53

 
26

4 
5.

10
 

0.
85

 
0.

05
 

3.
50

 
6.

70
 

F 
83

 
6.

50
 

0.
83

 
4.

94
 

8.
06

 
81

 
6.

19
 

0.
71

 
4.

86
 

7.
52

 
16

4 
6.

35
 

0.
78

 
0.

06
 

4.
88

 
7.

82
 

G
 

12
3 

7.
94

* 
1.

13
 

5.
82

 
10

.0
6 

10
4 

7.
54

 
0.

85
 

5.
94

 
9.

14
 

22
7 

7.
76

 
1.

03
 

0.
07

 
5.

82
 

9.
70

I2
 

C
 

28
 

3.
83

 
0.

52
 

2.
85

 
4.

81
 

15
 

3.
57

 
0.

49
 

2.
65

 
4.

49
 

43
 

3.
74

 
0.

52
 

0.
08

 
2.

76
 

4.
72

 
D

 
16

6 
4.

38
* 

0.
80

 
2.

88
 

5.
88

 
14

0 
4.

13
 

0.
70

 
2.

81
 

5.
45

 
30

6 
4.

26
 

0.
76

 
0.

04
 

2.
83

 
5.

69
 

E
 

12
2 

5.
86

* 
0.

88
 

4.
21

 
7.

51
 

12
7 

5.
58

 
0.

80
 

4.
08

 
7.

08
 

24
9 

5.
72

 
0.

85
 

0.
05

 
4.

12
 

7.
32

 
F 

95
 

7.
42

* 
1.

07
 

5.
41

 
9.

43
 

94
 

7.
00

 
0.

86
 

5.
38

 
8.

62
 

18
9 

7.
21

 
0.

99
 

0.
07

 
5.

35
 

9.
07

 
G

 
14

2 
8.

61
 

1.
23

 
6.

30
 

10
.9

2 
12

0 
8.

35
 

0.
92

 
6.

62
 

10
.0

8 
26

2 
8.

49
 

1.
10

 
0.

07
 

6.
12

 1
0.

56
C

 
C

 
14

3 
4.

16
 

0.
83

 
2.

60
 

5.
72

 
97

 
3.

93
 

0.
65

 
2.

71
 

5.
15

 
24

0 
4.

07
 

0.
77

 
0.

05
 

2.
62

 
5.

52
 

D
 

16
6 

5.
51

* 
0.

98
 

3.
67

 
7.

35
 

16
2 

5.
07

 
0.

99
 

3.
21

 
6.

93
 

32
8 

5.
29

 
1.

01
 

0.
06

 
3.

39
 

7.
19

 
E

 
16

6 
7.

51
* 

0.
99

 
5.

65
 

9.
32

 
13

1 
6.

91
 

0.
87

 
5.

27
 

8.
55

 
29

7 
7.

25
 

0.
98

 
0.

06
 

5.
41

 
9.

09
 

F 
22

2 
9.

90
* 

1.
31

 
7.

44
 

12
.3

6 
18

9 
8.

86
 

1.
05

 
6.

89
 

10
.8

3 
41

1 
9.

42
 

1.
30

 
0.

06
 

6.
98

 
11

.8
6

 
G

 
19

1 
12

.1
2*

 
1.

24
 

9.
79

 
14

.4
5 

14
9 

10
.9

1 
1.

24
 

8.
58

 
13

.2
4 

34
0 

11
.5

9 
1.

38
 

0.
07

 
9.

00
 1

4.
18

P
1 

B
 

40
 

3.
75

 
0.

57
 

2.
68

 
4.

82
 

39
 

3.
54

 
0.

47
 

2.
66

 
4.

42
 

79
 

3.
65

 
0.

53
 

0.
06

 
2.

65
 

4.
65

 
C

 
16

3 
4.

59
 

0.
82

 
3.

05
 

6.
13

 
15

0 
4.

46
 

0.
76

 
3.

03
 

5.
89

 
31

3 
4.

53
 

0.
79

 
0.

04
 

3.
04

 
6.

01
 

D
 

15
3 

6.
32

* 
0.

89
 

4.
65

 
7.

99
 

13
1 

6.
08

 
0.

99
 

4.
22

 
7.

94
 

28
4 

6.
21

 
0.

95
 

0.
06

 
4.

42
 

8.
00

 
E

 
16

2 
8.

21
* 

1.
06

 
6.

22
 

10
.2

0 
16

6 
7.

85
 

0.
92

 
6.

12
 

9.
58

 
32

8 
8.

02
 

1.
01

 
0.

06
 

6.
12

 
9.

92
 

F 
20

2 
10

.3
7*

 
1.

35
 

7.
83

 
12

.9
1 

16
0 

9.
83

 
1.

12
 

7.
72

 
11

.9
4 

36
2 

10
.1

3 
1.

28
 

0.
07

 
7.

72
 1

2.
54

 
G

 
15

1 
12

.2
5*

 
1.

27
 

9.
86

 
14

.6
4 

13
3 

11
.7

4 
1.

20
 

9.
48

 
14

.0
0 

28
4 

12
.0

1 
1.

26
 

0.
07

 
9.

64
 1

4.
38

P
2 

A
 

47
 

4.
07

 
0.

75
 

2.
66

 
5.

48
 

49
 

3.
80

 
0.

60
 

2.
67

 
4.

93
 

96
 

3.
93

 
0.

69
 

0.
07

 
2.

63
 

5.
23

 
B

 
84

 
4.

44
 

0.
72

 
3.

09
 

5.
79

 
71

 
4.

46
 

0.
86

 
2.

84
 

6.
08

 
15

5 
4.

45
 

0.
79

 
0.

06
 

2.
96

 
5.

94
 

C
 

14
4 

5.
79

* 
0.

99
 

3.
93

 
7.

65
 

11
6 

5.
44

 
0.

85
 

3.
84

 
7.

04
 

26
0 

5.
64

 
0.

95
 

0.
06

 
3.

85
 

7.
43

 
D

 
12

1 
7.

20
 

1.
11

 
5.

11
 

9.
29

 
13

3 
7.

15
 

1.
01

 
5.

25
 

9.
05

 
25

4 
7.

18
 

1.
06

 
0.

07
 

5.
19

 
9.

17
 

E
 

15
9 

9.
08

* 
1.

33
 

6.
58

 
11

.5
8 

14
3 

8.
60

 
1.

16
 

6.
42

 
10

.7
8 

30
2 

8.
85

 
1.

27
 

0.
07

 
6.

46
 

11
.2

4
 

F 
21

6 
11

.1
1*

 
1.

55
 

8.
20

 
14

.0
2 

17
2 

10
.6

8 
1.

26
 

8.
31

 
13

.0
5 

38
8 

10
.9

2 
1.

44
 

0.
07

 
8.

21
 1

3.
63

 
G

 
16

8 
13

.1
0*

 
1.

48
 

10
.3

2 
15

.8
8 

15
6 

12
.5

0 
1.

35
 

9.
96

 
15

.0
4 

32
4 

12
.8

1 
1.

45
 

0.
08

 
10

.0
8 

15
.5

4
M

1 
D

 
89

 
3.

94
 

1.
00

 
2.

06
 

5.
82

 
70

 
3.

68
 

0.
53

 
2.

68
 

4.
68

 
15

9 
3.

83
 

0.
83

 
0.

07
 

2.
27

 
5.

39
 

E
 

12
4 

4.
76

 
0.

73
 

3.
39

 
6.

13
 

11
1 

4.
53

 
0.

72
 

3.
18

 
5.

88
 

23
5 

4.
65

 
0.

73
 

0.
05

 
3.

28
 

6.
02

 
F 

11
3 

6.
20

* 
0.

88
 

4.
55

 
7.

85
 

10
0 

5.
78

 
0.

82
 

4.
24

 
7.

32
 

21
3 

6.
00

 
0.

88
 

0.
06

 
4.

35
 

7.
65

 
G

 
28

6 
8.

55
* 

1.
55

 
5.

64
 

11
.4

6 
27

2 
8.

14
 

1.
32

 
5.

66
 

10
.6

2 
55

8 
8.

35
 

1.
46

 
0.

06
 

5.
61

 
11

.0
9

M
2 

O
 

 
 

 
 

 
11

 
3.

45
 

0.
32

 
2.

85
 

4.
05

 
17

 
3.

61
 

0.
41

 
0.

10
 

2.
84

 
4.

38
 

A
 

43
 

3.
80

 
0.

43
 

2.
99

 
4.

61
 

26
 

3.
75

 
0.

44
 

2.
92

 
4.

58
 

69
 

3.
78

 
0.

43
 

0.
05

 
2.

97
 

4.
59

 
B

 
88

 
4.

56
 

0.
65

 
3.

34
 

5.
78

 
88

 
4.

39
 

0.
83

 
2.

83
 

5.
95

 
17

6 
4.

48
 

0.
74

 
0.

06
 

3.
09

 
5.

87
 

C
 

16
2 

5.
93

 
0.

95
 

4.
14

 
7.

72
 

13
3 

5.
68

 
0.

88
 

4.
03

 
7.

33
 

29
5 

5.
82

 
0.

93
 

0.
05

 
4.

07
 

7.
57

 
D

 
16

6 
7.

80
 

1.
19

 
5.

56
 

10
.0

4 
17

6 
7.

53
 

1.
05

 
5.

56
 

9.
50

 
34

2 
7.

66
 

1.
13

 
0.

06
 

5.
54

 
9.

78
 

E
 

16
7 

9.
74

 
1.

28
 

7.
33

 
12

.1
5 

13
7 

9.
38

 
1.

13
 

7.
26

 
11

.5
0 

30
4 

9.
58

 
1.

22
 

0.
07

 
7.

29
 

11
.8

7
 

F 
13

5 
11

.5
6*

 
1.

14
 

9.
42

 
13

.7
0 

12
0 

10
.9

8 
1.

05
 

9.
01

 
12

.9
5 

25
5 

11
.2

9 
1.

13
 

0.
07

 
9.

17
 1

3.
41

 
G

 
26

0 
13

.6
3*

 
1.

45
 

10
.9

0 
16

.3
6 

26
4 

13
.2

4 
1.

48
 1

0.
46

 
16

.0
2 

52
4 

13
.4

4 
1.

48
 

0.
06

 
10

.6
6 

16
.2

2

T
A

B
L

E
 4

. A
ve

ra
ge

 a
ge

 o
f c

hi
ld

re
n

 w
it

hi
n

 a
 s

ta
ge

 fr
om

 a
 u

n
if

or
m

 a
ge

 d
is

tr
ib

u
ti

on
 o

f 8
7 

m
al

es
 a

n
d 

84
 fe

m
al

es
 fo

r 
ea

ch
 y

ea
r 

of
 a

ge
 3

-1
6.

 U
n

de
rs

co
re

 m
ea

n
s 

ag
e 

n
ot

 
si

gn
ifi

ca
n

tl
y 

di
ff

er
en

t 
be

tw
ee

n
 s

ex
es

. A
st

er
is

k 
m

ea
n

s 
ag

e 
si

gn
ifi

ca
n

tl
y 

di
ff

er
en

t 
(P

 <
 0

.0
1)

 b
et

w
ee

n
 s

ex
es

.

DEMIRJIAN TOOTH FORMATION STAGES



24

Liversidge HM. 2003.  Worldwide variation in human dental 
development. In: Thompson JL, Nelson A, Krovitz G, 
editors. Growth and development in the genus Homo. 
Cambridge: Cambridge University Press. p 73-113.

Liversidge HM. 2008. Dental age revisited. In: Irish JD, 
Nelson GC, editors. Technique and application in dental 
anthropology. Cambridge:  Cambridge University  
Press.  p 234-265.

Liversidge HM. 2009. Permanent tooth formation as a 
method of estimating age. Front Oral Biol 13:153-157.

Liversidge HM, Chaillet N, Mörnstad H, Nyström M, 
Rowlings K, Taylor J, Willems G. 2006. Timing of 
Demirjian tooth formation stages. Ann Hum Biol 33: 
454-470.

Liversidge HM, Smith BH, Maber M. 2010. Bias and 
accuracy of age estimation using developing teeth 
in 946 children. Am J Phys Anthropol (doi:10.1002/
ajpa.21349).

Liversidge HM, Speechly T. 2001. Growth of permanent 
mandibular teeth of British children aged 4-9 years. 
Ann Hum Biol 28:256-262.

Liversidge HM, Townsend G. 2006. Tooth formation in 
Australian Aborigines. In: Zadzinska E, editor. Current 
trends in dental morphology research.  Lodz: University 
of Lodz Press. p  405-410.  

Maber M, Liversidge HM, Hector MP. 2006. Accuracy of age 
estimation of radiographic methods using developing 
teeth. Forensic Sci Int 159:S68-73.

McKenna CJ, James H, Taylor JA, Townsend GC. 2002. 
Tooth development standards for South Australia. Aust 
Dent J 47:223–227.

Moananui RT, Kieser JA, Herbison P, Liversidge HM. 2008. 
Advanced dental maturation in New Zealand Maori 
and Pacific Island children. Am J Hum Biol 20:43-50.

Moorrees CFA, Fanning EA, Hunt EE. 1963. Age variation 
of formation stages for ten permanent teeth. J Dent Res 
42:1490-1502.

Nolla CM. 1960. The development of the permanent teeth. 
J Dent Child 27:254–266.

Nyström M, Haataja J, Kataja M, Evalahti M, Peck L, 
Kleemola–Kujala E. 1986. Dental maturity in Finnish 
children, estimated from the development of seven 
permanent mandibular teeth. Acta Odontol Scand 
44:193–198.

Nyström M, Ranta R, Kataja M, Silvola H. 1988. Comparisons 
of dental maturity between the rural community of 
Kuhmo in north eastern Finland and the city of Helsinki. 
Commun Dent Oral Epidemiol 16:215–217.

Reid DJ, Dean MC. 2006. Variation in modern human 
enamel formation times. J Hum Evol 50:329-346.

Smith BH. 1991. Standards of human tooth formation 
and dental age assessment.  In: Kelly MA, Larsen CS, 
editors. Advances in dental anthropology. New York: 
Wiley-Liss. p 143-168.

Taranger J. 1976. Evaluation of biological maturation by 
means of maturity criteria. Acta Paediatr Scand 258, 
suppl, 77-82.

Teivens A, Mörnstad H. 2001. A comparison between dental 
maturity rate in the Swedish and Korean populations 
using a modified Demirjian method. J Forensic 
Odontostomatol 19:31–35.

Willems G, Van Olmen A, Spiessens B, Carels C. 2001. 
Dental age estimation in Belgian children: Demirjian’s 
technique revisited. J Forensic Sci 46:893–895.

Fig. 7. Preliminary results of maturation of stage H of the mandibular first molar from world regions.  Left: smoothed 
curves (proportion of girls vs. age). Right: 95% confidence interval of mean age for Australian Aborigine, Inuit, Japanese,  
Maori and Pacific Islanders, South African Black, South African Cape Coloured,  UK Bangladeshi, UK White girls.

H.M. LIVERSIDGE.