Sub-Adult Age from Dental Eruption and Epiphyseal Fusion

Francis Moorrees, Elizabeth Fanning, and Eugene Hunt published their landmark developmental atlas in the American Journal of Physical Anthropology in 1963, based on longitudinal radiographic records of American children from the Fels Longitudinal Study in Ohio. Their fundamental contribution was formalising dental development as a series of thirteen stages that could be reliably scored from periapical radiographs by any trained observer. The stages run from initial cusp formation (Ci) through crown completion (Cc), then through root initiation (Ri), root formation (R one-quarter, R one-half, R three-quarters), root completion (Rc), apex closure (A one-half, Ac), and for multi-rooted teeth, the separate staging of furcation and root trunk length.

The Moorrees-Fanning-Hunt framework was applied to ten mandibular teeth, and age means and standard deviations were derived for males and females separately. The system's strength was its precision: by scoring multiple teeth simultaneously and comparing the observed profile against the sex-specific reference charts, a forensic anthropologist could generate an age estimate with a standard deviation of roughly six to twelve months for children in the five to thirteen year range. For the third molar, which initiates later and whose stages fall across the legally critical fourteen to twenty-one year window, the standard deviation widened considerably.

The Moorrees-Fanning-Hunt framework was extended by other researchers to maxillary teeth and to populations outside North America. Its staged scoring approach was directly adopted by Demirjian ten years later, and several Indian studies referenced in the ICMR standards use a modified Moorrees staging scheme applied to North Indian and South Indian reference samples.

One practical limitation of the 1963 atlas is that it requires the observer to interpolate graphically from printed age-versus-stage charts, which introduced inter-observer variability. Modern implementations use the regression equations derived from the underlying Fels data, which allow a direct numerical age prediction from staged tooth scores.

Arto Demirjian, Harold Goldstein, and John Tanner published their dental maturity method in Human Biology in 1973, based on 1,446 radiographs of French-Canadian children aged two to nineteen years from Montreal. The method's defining innovation was a seven-stage scoring scheme (stages A through G, with stage H added for third molars in later revisions) applied to seven mandibular teeth on the left side: the central incisor, lateral incisor, canine, first premolar, second premolar, first molar, and second molar. Each tooth is scored A through G from a set of illustrated definitions; the stages have crisp visual criteria that minimise inter-observer variation.

Each tooth-stage combination is assigned a weighted numerical score from a sex-specific table. The scores for all seven teeth are summed to give a total dental maturity score between 0 and 100, which is then converted to a dental age using a conversion table derived from the Montreal reference sample. The maturity-score approach is powerful because it separates the measurement (scoring teeth against illustrated criteria) from the interpretation (converting the score to an age). If a new reference sample is collected from a different population, only the score-to-age conversion table needs to be recomputed; the staging criteria and the individual tooth weights remain constant. This portability drove widespread international adoption: by the late 2000s, Demirjian scores had been validated against reference samples from Belgium, France, Spain, Turkey, Brazil, Japan, South Korea, India, and dozens of other countries.

The Indian validation data are particularly important for practitioners working in South Asian contexts. Chaillet and Willems published a pooled international reference set in 2004. For Indian populations specifically, several studies published between 1995 and 2015 found that the Demirjian method overestimates dental age by six to eighteen months for Indian children, a systematic bias attributable to earlier dental maturation in Indian compared to French-Canadian reference children. The forensic implication is that applying the original Demirjian tables without population correction to an Indian sub-adult will generate an age estimate that is slightly too high. Indian practitioners should apply the population-specific correction factors or use the AlQahtani London Atlas, which was constructed with a broader international base.

Sakher AlQahtani, Mike Hector, and Helen Liversidge published the London Atlas of Human Tooth Development and Eruption in the American Journal of Physical Anthropology in 2010. The atlas was constructed from 704 individuals from the collection held at the Odontological Museum of the Royal College of Surgeons of England, a collection built on documented post-mortem records from individuals who died in London between 1826 and 1996 and whose ages at death are known from parish and civil records. The collection includes individuals of British, European, African, and South Asian ancestry, which gives the London Atlas considerably more ethnic breadth than the Moorrees (white American) and Demirjian (French-Canadian) reference samples.

The atlas covers the full developmental lifespan from birth to 23.5 years, using illustrated plates that show both the deciduous and permanent dentitions simultaneously for each age interval. Each plate depicts the expected stage of crown and root development for every tooth, the eruption status (unerupted, partial eruption, fully erupted), and the pattern of deciduous tooth resorption. The illustrations were derived from photographs of sectioned, macerated teeth from the collection, not from radiographs, which gives them a three-dimensional anatomical clarity that earlier atlases could not match.

The matching procedure in casework is direct: the examiner arranges available dental evidence (either from radiographs or from the dry remains) to match the plate most closely. The matching plate carries a stated age range at the top. AlQahtani et al. reported an accuracy of plus or minus one year for ages up to 16, widening to plus or minus 2.5 years for the 16 to 22 range where third molar variability dominates.

The London Atlas is the current standard in INTERPOL DVI dental age estimation, adopted by the UK Crown Prosecution Service guidance on child age estimation, used by the International Olympic Committee's Medical Commission for age verification in sport, and referenced in the US Department of Homeland Security forensic age assessment protocols. In India, the atlas is used by several AIIMS Forensic Medicine departments and has been evaluated against an Indian cohort by Kumar et al. (2014, Journal of Forensic and Legal Medicine) with reported mean absolute errors of approximately 0.8 to 1.3 years across age groups.

A consistent biological finding across all major reference populations is that mandibular teeth erupt slightly earlier than their maxillary counterparts, with the exception of the maxillary canine, which erupts before the mandibular canine in most populations. The magnitude of the mandibular-first advantage is approximately three to six months for most tooth types, though the canine reversal is more pronounced.

In dry skeletal remains where the maxilla is damaged or missing, the forensic anthropologist must rely on the mandibular eruption sequence alone, which is acceptable given that the mandibular sequence is the one most densely represented in reference databases. In living person examination where oral access is limited, knowing the arch-specific sequence allows the examiner to extract maximum information from whatever teeth are accessible.

The clinically and forensically important eruption milestones in the permanent dentition are: first molar eruption at approximately 5.5 to 7 years (the first permanent tooth to appear, in both arches); central incisor eruption at 6 to 8 years; canine eruption in the mandible at 9 to 11 years; second premolar eruption at 10 to 13 years; and second molar eruption at 10 to 14 years. Third molar eruption onset ranges from 16 to 22 years, with substantial individual and population variation.

The third molar deserves special attention because it sits squarely across the legal threshold of majority in most jurisdictions: 18 years in the United Kingdom, European Union, and India; 18 years under the US federal juvenile justice framework (though individual states vary). The third molar's root development stages can, under optimal conditions, distinguish a person who is almost certainly under 18 from one who is almost certainly over 18, but the wide variability (including 25 to 35 per cent of individuals who congenitally lack one or more third molars) means that the third molar alone is insufficient and must be combined with other methods.

The Scheuer-Black framework, systematised in Louise Scheuer and Sue Black's definitive textbook "Developmental Juvenile Osteology" (Academic Press, 2000; second edition by Scheuer, Black, and Alison Christie, 2007), catalogues the fusion timing of every epiphysis in the human skeleton from foetal age through early adulthood. The framework integrates radiographic studies, autopsy series, and population-specific data to provide mean ages and ranges for each fusion event.

Epiphyseal fusion occurs when the growth plate cartilage between the epiphysis (the end of the bone) and the diaphysis (the shaft) ossifies and is replaced by bone, eliminating the growth plate and completing the bone's longitudinal growth. The fusion sequence is broadly predictable and is used forensically to define age brackets rather than point estimates: an epiphysis that shows no sign of fusion indicates the individual is probably below the minimum of the fusion age range; an epiphysis that shows complete fusion indicates the individual has reached or passed the maximum of the range.

The Scheuer-Black framework recognises five observable fusion stages: unfused, beginning to fuse (less than one-third fused), actively fusing (one-third to two-thirds fused), almost fused (two-thirds to nearly complete), and completely fused. The transition from unfused to completely fused typically spans two to four years for any individual epiphysis. The critical implication for age estimation is that the most useful epiphyses for a given case are those for which the transition period straddles the age of interest.

The earliest fusions in practical forensic use involve the distal humeral epiphyses (medial epicondyle fusion 14 to 18 years in females, 15 to 20 in males) and the distal radius (fusion begins around 14 years in females, 15 in males; complete by 18 in females, 19 in males). The latest fusions involve the medial clavicle (see Section 7) and the sacral segments, which extend to the late twenties.

The Indian reference data for epiphyseal fusion deserve explicit mention because population differences are real. Greulich and Pyle's classic American atlas (1959) is still widely used in Indian medicolegal practice, but several Indian studies, including those published in the Journal of Anatomy and the Journal of Forensic Medicine and Toxicology, have shown that Indian populations achieve certain fusion milestones slightly earlier than the Greulich-Pyle reference sample would predict. The Forensic Medicine departments at AIIMS New Delhi, AIIMS Bhubaneswar, and several state medical colleges have contributed to a developing body of Indian fusion data, though a comprehensive national atlas comparable to Scheuer-Black for Indian populations remains a gap as of 2025.

The medial clavicular epiphysis (the sternal end of the clavicle) typically does not begin fusing until age 16 to 18 years, continues fusing through the early twenties, and reaches complete fusion on average between 21 and 25 years, with some individuals showing incomplete fusion as late as 30. This timeline makes the medial clavicle uniquely informative for the forensic question that most frequently arises in living person age assessment: is this individual at least 18 (or 21) years old?

The forensic significance of this epiphysis was formalised by Andreas Schmeling and Axel Olze at the Charite Institute for Legal Medicine in Berlin, who published a systematic series of studies beginning in 2003 establishing a four-stage scoring system for medial clavicular fusion visible on computed tomography (CT):

Stage 1: epiphysis not yet ossified. Stage 2: epiphysis ossified but not fused to the clavicular shaft. Stage 3: partial fusion of the epiphysis to the shaft (the fusion line is visible but not complete). Stage 4: complete fusion (no visible fusion line).

Schmeling et al.'s 2004 and 2008 papers reported that complete fusion (Stage 4) is essentially never seen before age 21 in either sex. Partial fusion (Stage 3) begins to appear around age 17 in females and 18 in males. The absence of any fusion (Stage 1-2) makes it highly unlikely, though not impossible, that the person is over 20.

The CT-based method requires access to cross-sectional imaging, which is standard in German, British, French, and Dutch forensic age assessment protocols for living persons. In India, CT availability in forensic contexts has improved substantially since 2015, and several major AIIMS centres and state FSLs can now conduct medial clavicle CT assessment. The INTERPOL DVI Standing Committee recommends medial clavicle CT as a component of the multi-method protocol for both living-person and skeletal age assessment.

The forensically most important fusion events, ordered by the age window they address:

Distal radius and distal ulna: fusion complete in most females by 17 to 18 years, in most males by 18 to 19 years. An unfused distal radius in a female strongly suggests she is under 17; complete fusion does not rule out ages below 18.

Iliac crest: the iliac crest apophysis fuses between 15 and 23 years, with the posterior iliac crest completing slightly before the anterior. It is visible on plain radiographs and is one of the most reliable indicators for the 16 to 20 age window.

Sacral segments S1 and S2: the S1 and S2 vertebral epiphyses show active fusion between approximately 18 and 25 years. Incomplete S1-S2 fusion combined with a partially fused medial clavicle strongly suggests an age below 25. The sacral fusion sequence is assessed from CT images; plain radiographs are generally inadequate.

Medial clavicle: complete fusion after approximately 21 to 25 years, as described in Section 7. The combination of the medial clavicle with the wisdom tooth root stage and, where available, the wrist radiograph (which captures the distal radius, distal ulna, and carpal development) forms the Schmeling tri-modal protocol.

Basilar (spheno-occipital) synchondrosis: this junction between the sphenoid body and the basiocciput closes on average between 18 and 25 years, with females completing earlier than males on average. It is visible on the base of the skull in dry skeletal remains and on lateral skull radiographs. Partial closure suggests an age under 22 to 25; complete closure is consistent with but does not require an age over 18.

The Schmeling et al. 2008 framework, published in the International Journal of Legal Medicine and adopted as the standard protocol by the Study Group on Forensic Age Diagnostics (AGFAD) in Germany and referenced by equivalent bodies in Austria, Switzerland, the Netherlands, Spain, and the United Kingdom, specifies a four-component assessment for living persons of uncertain age.

Component one is a physical examination documenting anthropometric measurements, secondary sexual development (Tanner staging), and any pathological findings that might affect skeletal maturation. Component two is a dental examination including dental radiographs (orthopantomogram or periapical films), staged using the current standard atlas (AlQahtani preferred for international populations). Component three is a left-hand-and-wrist radiograph, documenting the fusion status of the distal radius, distal ulna, and carpal bones, compared against appropriate population-specific reference data. Component four, recommended when the question is whether the person is above or below a specific threshold (typically 18 or 21 years), is a CT examination of the medial clavicle for fusion staging.

The protocol requires that all four components be assessed independently and that the findings from each component be combined probabilistically to generate a final age estimate expressed as a range with a confidence interval. The protocol explicitly forbids reporting a single-year point estimate; the minimum reporting standard is a range that reflects the full uncertainty of the assessment.

In the United Kingdom, the Home Office and the Association of Chief Police Officers (now the National Police Chiefs' Council) reference the Schmeling protocol in guidance on age assessment for asylum seekers. The Independent Age Assessment framework developed by local authority social services in England operates somewhat differently, relying primarily on a holistic assessment by social workers, but medical evidence including dental radiographs and wrist radiographs can be admitted as supporting evidence. In India, the Juvenile Justice Act 2015 and the JJ Model Rules 2016 specify that age determination follows a cascading process: birth certificate first, school records second, and finally a medical board examination including "bone ossification test" and dental examination. The JJ Act framework does not specify a particular atlas or fusion scale, creating variability in practice across states.

The United States Department of Homeland Security uses a similar multi-modal approach for age verification of unaccompanied minors; the specific radiographic standards applied vary by contracting laboratory, but the foundational principles follow the Schmeling framework.

The Bagga 2008 dental atlas, published by Anjana Bagga and colleagues at the Post Graduate Institute of Medical Education and Research (PGIMER), Chandigarh, and the dental development standards produced under Indian Council of Medical Research (ICMR) multi-centre studies, constitute the principal India-specific reference data for dental age estimation in the sub-adult range.

Bagga's atlas covered Punjabi children aged 3 to 16 years, using orthopantomographic radiographs and applying both Demirjian staging and eruption-based assessment. The key finding, consistent with earlier Indian studies, was that Indian children show earlier dental maturation compared to the original Demirjian French-Canadian reference: the mean overestimation when applying the original Demirjian score-to-age conversion to North Indian children was approximately 0.8 to 1.4 years across the age range, with the greatest bias in the 8 to 12 year window. The practical correction is to subtract approximately one year from the Demirjian-predicted age when the individual is from a North Indian population and the original conversion tables are used.

The ICMR standards, published as part of the "Growth and Physical Development of Indian Children" series and summarised in the ICMR Reports on Child Health 2010 and 2018, provide population-level benchmarks for dental eruption timing that are relevant for comparison when the forensic question involves a child from a rural or nutritionally diverse background. These standards acknowledge regional variation within India: South Indian populations show slightly different mean eruption ages compared to North Indian populations, and tribal populations show additional variation.

For the medial clavicle and epiphyseal fusion data, Indian-specific published studies are sparser. The most cited fusion studies for Indian populations come from the Forensic Medicine and Toxicology departments of AIIMS New Delhi, Bombay (Mumbai), and Madras (Chennai), published across several decades in the Journal of Forensic Medicine and Toxicology and Medico-Legal Journal of India. These studies generally support the Scheuer-Black ranges for major epiphyses but report somewhat earlier fusion for the medial clavicle compared to the European studies on which the Schmeling thresholds are based, a finding that, if confirmed by larger Indian samples, would reduce the confidence with which a partially fused medial clavicle can be used to exclude adulthood in an Indian population.