Sex Estimation from Teeth and Dental Dimensions

Crown size is determined during odontogenesis, the window of tooth development before eruption. The biochemical environment during that window, including circulating sex hormones, influences the final dimensions. Males undergo a longer crown mineralisation period for several tooth classes, which correlates with larger absolute crown size. This effect is modest, rarely more than a few tenths of a millimetre in mean mesiodistal diameter, but it is consistent across populations even if the magnitude varies.

The canines show the strongest dimorphism in the permanent dentition, a pattern shared with other hominids and thought to reflect sexual selection pressures far back in primate evolution. The mandibular canine in particular has been the most studied tooth for sex estimation in forensic odontology. Other teeth show dimorphism too, but weaker; combining measurements from multiple teeth via discriminant function analysis squeezes out more signal than any single tooth provides.

Rao, Rao, Pai, and Kotian described the mandibular canine index (MCI) in 1989 as a practical single-tooth metric. The formula is straightforward: measure the mesiodistal crown width of the left or right lower canine (whichever is present), divide by the intercanine distance measured at the canine tips, and multiply by 100. Male canines are relatively wider, so males tend to produce higher MCI values. A threshold is calculated from a reference sample, and specimens above the threshold are classed as male, below as female.

Published accuracy ranges from roughly 70 to 85 percent, depending on the study population. Indian studies have repeatedly found MCI useful, with most reporting accuracy in the 75 to 85 percent range. Studies on European and American samples find similar but not identical thresholds, confirming that the ratio is not universal. A threshold derived from one population applied directly to another will reduce accuracy, sometimes substantially.

The practical limitation is that MCI is a single-tooth method in a mouth with 32 teeth. When only the mandibular canine is present, it is the best available measure. When more teeth survive, a multivariate approach will outperform it. Analysts should report MCI as one input among several rather than as the sole basis for a sex determination.

Standard dental odontometrics measures the mesiodistal (front-to-back) and buccolingual (cheek-to-tongue) crown diameters for each tooth class. Males average larger values for both dimensions across most teeth, but the degree of overlap between male and female distributions is substantial for any single tooth. Discriminant function analysis addresses this by combining the measurements, each weighted by its discriminating power, into a single equation whose output score is compared against a threshold to classify the unknown.

• Which teeth to measure: canines, premolars, and first molars together give better results than any one class alone. Anterior teeth are particularly useful because they preserve well and show consistent dimorphism.
• Which dimensions: mesiodistal diameter is more dimorphic and less subject to wear-related reduction than buccolingual diameter; both are included in full analyses.
• Reference population matching: a DFA equation trained on a Japanese sample will produce a different function than one trained on a Nigerian or a Brazilian sample. Applying the wrong equation inflates error rates.
• Wear and fragmentation: significant occlusal wear reduces crown height and can erode enamel breadths. Moderate or heavy wear may make measurements unreliable; the analyst should document the wear grade.

When a complete or near-complete permanent dentition is available and measured with a population-matched DFA, accuracy in controlled studies frequently reaches 80 to 90 percent. Real casework, where only some teeth survive and the population of origin is itself uncertain, tends to lower accuracy. The honest position is to report a probabilistic estimate with the method and reference source named, not a categorical sex assignment that the data do not support.

In XX individuals, one of the two X chromosomes is transcriptionally silenced and condenses into a compact chromatin mass called the Barr body (after Murray Barr, who described it in 1949 in nerve cell nuclei). In dental pulp cells this body appears as a small dark speck pressed against the nuclear envelope under a standard haematoxylin-and-eosin stain. In polymorphonuclear leukocytes trapped in the pulp, a related condensed X structure forms a drumstick-shaped appendage on the nuclear lobe.

The forensic appeal is that these structures do not require DNA extraction: a thin section of pulp tissue, stained and mounted, is all that is needed. The limitation is tissue quality. Barr bodies require relatively intact nuclei, which means fresh or recently extracted teeth, well-preserved embalmed tissue, or teeth stored under dry conditions for a relatively short time. In skeletonised remains or teeth that have been buried for decades, nuclear architecture degrades and the technique becomes unreliable.

Amelogenin is the dominant protein of forming dental enamel. The gene encoding it sits on both sex chromosomes: AMELX on the X chromosome and AMELY on the Y chromosome. A short intron within the gene differs in size between the two versions. Standard PCR primers flanking this intron amplify a fragment of roughly 106 base pairs from AMELX and about 112 base pairs from AMELY in most human populations. Agarose gel electrophoresis or capillary electrophoresis then shows a single band for XX individuals (only the X product) and two bands for XY individuals.

The method's main advantages are sensitivity and substrate versatility. Dentine is one of the best sources of ancient DNA in the human body: the mineral matrix shields the molecule from oxidative damage, and pulp tissue, even after years in soil, can yield amplifiable DNA. Amelogenin amplicons are short, making them recoverable even when DNA is highly fragmented. The assay is a standard component of forensic STR kits (AmpFlSTR, PowerPlex) so it integrates naturally into a standard DNA profiling workflow.

The caveats are worth knowing. First, a false female call occurs when AMELY is deleted. AMELY deletion has been documented at frequencies between 1 in 60 and 1 in 800 in different populations, being most common in some South and Southeast Asian populations. An XY individual with AMELY deletion shows only the X band and is misclassified as female. Second, contamination with female contributor DNA can mask a male profile at low template concentrations. Third, in the rare case of 45,X (Turner syndrome), no Y band is expected; in 47,XXY (Klinefelter syndrome), both bands appear, as in XY.

A practical hierarchy runs from molecular to morphometric to microscopic. When DNA quality is adequate, amelogenin gives the most reliable sex call, but a positive molecular result is strengthened when the odontometric evidence also leans the same way. When DNA is too degraded to amplify, morphometrics and possibly Barr body examination become the only tools available, and the conclusion should reflect that the method's accuracy ceiling is lower.

1. Step 1: Assess preservation
   Inspect the teeth for gross completeness, wear grade, and any signs of thermal or chemical damage. Note how many teeth are present and whether pulp chambers are intact. This determines which methods are feasible.
2. Step 2: Morphometric analysis
   Measure available crown diameters with a digital calliper to 0.1 mm. Calculate MCI if the mandibular canine is present. Apply a DFA matched to the probable population of the decedent and record the discriminant score and the posterior probability.
3. Step 3: Molecular analysis
   Extract DNA from dentine (surface-cleaned to remove exogenous contamination), quantify, and amplify the amelogenin locus. Interpret the band pattern and note any ambiguity. If a full STR kit was run, the amelogenin result is already in the output.
4. Step 4: Microscopic analysis (where tissue quality permits)
   Section fresh or well-preserved pulp tissue, stain with haematoxylin and eosin, and search for Barr bodies or drumstick appendages in at least 500 cells. Report percentage of positive cells and compare to published reference ranges.
5. Step 5: Report synthesis
   Combine all lines of evidence into a single conclusion with an explicit confidence statement. If all three methods agree, confidence is high. If molecular and morphometric disagree, investigate AMELY deletion or measurement error before finalising. Never exceed the accuracy the weakest-method evidence supports.

An important note on population-specific norms: the accuracy figures cited for MCI, DFA, and Barr body studies come largely from Indian, East Asian, and European reference samples. When the population of origin is unknown, as in unidentified remains from a mass casualty or an international disaster, the analyst should apply whichever reference populations are most plausible given the recovery context and acknowledge the added uncertainty.