Adult Age: Gustafson, Kvaal, Cementum, Racemisation

Gösta Gustafson published his method in the Journal of the American Dental Association in 1950. He sectioned 41 extracted teeth from individuals of known age and scored six visible degenerative changes on each section, assigning each a score of 0 (absent), 1 (slight), 2 (moderate), or 3 (advanced). The six criteria are attrition (A), periodontosis (P), secondary dentine (S), cementum apposition (C), root resorption (R), and root transparency (T).

• Attrition (A): wear of the incisal or occlusal surface. Highly variable because it depends on diet, occlusion, and parafunctional habits such as bruxism. One of the weaker single predictors.
• Periodontosis (P): apical migration of the gingival attachment and bone loss. Also influenced by oral hygiene and systemic disease, so variable in the population.
• Secondary dentine (S): progressive reduction of pulp chamber volume. One of the more consistent indicators across individuals, scored on the sectioned tooth.
• Cementum apposition (C): thickening of cementum particularly at the root apex. Scored from the section but best quantified by ring counting.
• Root resorption (R): pitting and irregular loss of root substance at the apex. Scores 0 in young teeth; increases in older individuals.
• Root transparency (T): advancing translucency of the root dentine from the apex. One of Gustafson's most reliable criteria, relatively independent of lifestyle factors.

Gustafson's original regression gave a standard error of about 3.6 years, which was encouraging. Subsequent studies applying his method to independent samples found standard errors of 8-10 years, revealing that the method was optimistic when transferred to different populations. Later researchers (Bang and Ramm 1970, Maples 1978, and others) applied the method to larger and more varied samples, confirmed root transparency as the single most reliable indicator, and proposed modified regression equations. Root transparency still features in most modern multi-criteria schemes.

Sigrid Kvaal and colleagues published a non-destructive approach in 1995 based on the observation that the pulp cavity shrinks predictably with age as secondary dentine is deposited. They defined measurement ratios on periapical radiographs: the ratio of pulp length to root length, and the ratio of pulp width to root width at three defined points along the root. These ratios were combined in a multiple regression equation to estimate age.

The method covers six tooth types (maxillary and mandibular central incisors, lateral incisors, second premolars, and mandibular first premolars) and uses sex-specific regression coefficients. Standard errors in the original study ranged from 8.6 to 11.5 years, which is wider than Gustafson's optimistic original but consistent with validated multi-method adult estimates. The main advantage is that no tooth extraction is needed: a set of periapical films or a cone-beam CT (CBCT) scan, which gives volumetric pulp measurement, is sufficient.

The cementum that covers the root surface is not deposited in one uniform layer. It forms in seasonal increments, creating alternating translucent and opaque bands visible in ground sections viewed under transmitted polarised light. One light-dark band pair is deposited per year under normal conditions, making the count a direct annual record. Counting the rings from the point of tooth eruption and adding the eruption age gives a total age estimate.

Cementum annulation was described as a forensic age tool by Gustafson (1950) and systematically validated by Stott, Sis, and Levy (1982), among others. Studies typically report accuracy within two to three years when sections are prepared and read under optimal conditions. The method is destructive (a tooth must be extracted and sectioned) and technically demanding. Observer variability in ring counting, and confusion between true annulations and artefact lines from processing, are the main sources of error.

• The single-rooted teeth (canines, incisors, first premolars) give the best sections because the cementum is thickest at the apex and the ring geometry is simplest.
• Disease, metabolic disruption, and certain medications can create false lines (arrest lines) that can be mistaken for annual rings and inflate the count.
• Cross-polarised light and digital image enhancement have improved discrimination of true annual rings from processing artefacts in modern practice.

The amino acids in living organisms are almost exclusively in the L-form (left-handed). After proteins are synthesised, a slow spontaneous chemical reaction converts some L-amino acids to their D-form (right-handed) mirror images. This process is called racemisation. The rate constant for aspartic acid, the most reactive natural amino acid, is well-characterised and is minimally affected by normal body temperature variation. Because tooth enamel does not remodel after it is first laid down, the D/L aspartic acid ratio in enamel reflects the time elapsed since enamel formation. From that ratio and the known rate constant, an age can be calculated.

Helfman and Bada published the first forensic application in 1975, reporting aspartic acid racemisation in tooth enamel from living humans (Proceedings of the National Academy of Sciences USA, 72:2891-2894); a follow-up paper on dentine appeared in Nature in 1976. Subsequent work, including large validation studies by Ritz-Timme and colleagues, confirmed standard errors of approximately two to three years under controlled laboratory conditions. This is substantially tighter than any morphological adult method. The main practical limitation is that the analysis requires laboratory chemistry (hydrolysis of the enamel protein followed by high-performance liquid chromatography or gas chromatography to separate D- and L-amino acids), and the tooth enamel is consumed in the process.

In a casework setting, adult dental age is rarely determined from a single method. Best practice combines at least two independent approaches: typically a morphological multi-criteria assessment (Gustafson, or a validated modification) alongside either cementum ring counting or, where laboratory access exists, aspartic acid racemisation. The two estimates are then reviewed for consistency and combined into a reported range.

Even with the best available chemistry, a five-year confidence interval around the estimate is realistic for most adult cases. This has direct implications for legal and administrative uses: a method that gives ±3 years at the 95% confidence level cannot reliably separate a 16-year-old from an 18-year-old when the point estimate is 17. Any report used for an age-threshold decision should explicitly state this uncertainty, not present a point estimate as a definitive birthdate.