Tooth crown diameters, such as maximum mesiodistal and buccolingual measurements, have frequently been the subject of analysis by dentists, geneticists, and physical anthropologists. Crown diameters are, however, the composites of dentine, a neural crest (ectomesenchymal) derivative, and the enamel covering, which is an ectodermal tissue, and the dimensions of these two tissues need not co-vary.

There are few data for enamel thickness in humans. Physical anthropologists have quantified enamel thickness in various non-human primates, but none tested for sexual dimorphism. Often-cited studies of humans, based on direct measure ments of sectioned teeth, include works by Fejerskov et al. (1973), Gillings and Buonocore (1961a) and Gillings and Buonocore (1961b) and Shillingsburg and Grace (1973). These researchers grouped sexes and, occasionally, tooth types as well. The same holds for more recent studies of humans, where attention has focused on comparisons of thicknesses among different morphological crown structures. Perhaps the work of Stroud et al. (1994) is the only study where sex and ethnic origin were both controlled.

Moss and Moss-Salentijn (1977) speculated that canines—the most sexually dimorphic teeth in humans—are bigger in males than females because of greater enamel thickness in males, but this was speculative in the absence of relevant data. On the other hand, it does appear that aneuploidy significantly influences enamel thickness. Alvesalo and Tammisalo (1981) and Alvesalo and Tammisalo (1985) showed that the absence of a second sex chromosome (i.e., 45,XO) resulted in substantially thinner enamel, but the size of the dentine component of the crown was unaffected. Conversely, males with Klinefelter syndrome (47,XXY) have much thicker enamel. It also is evident that some Mendelian traits have an overt influence on enamel thickness, which sheds light on the genetic regulation of enamel thickness and structure.

Our purpose now was to describe the mesiodistal crown and enamel thicknesses of the four permanent maxillary incisors. Specifically we sought to localize the source of sexual dimorphism in crown size to the dentine, the enamel, or a combination of the two, and to assess the strengths of statistical associations between enamel and dentine thicknesses, as these two tissues are active at different times during a tooth’s formation. The study should also provide normative data against which samples with genetic, chromosomal, or other abnormalities affecting dentine and enamel formation can be compared quantitatively.

Standardized periapical radiographs were obtained from adolescents with intact, caries-free, macroscopically unworn maxillary incisors. The sample (n=115) consisted of American caucasoids between the ages of 13 and 17 years. Orientation of the X-ray source and the radiograph was controlled with a parallelling apparatus, and a strip of adhesive material containing a radio-dense millimetric grid (XRG, Ada Products, Milwaukee, WI, USA) was placed so that it was imaged on the radiograph but did not overlie the tooth itself. This grid provided a check of radiographic enlargement.

Data were obtained from a longitudinal clinical trial conducted on routine dental patients for other purposes, so multiple films were available for each incisor on each individual. However, as none of the cases had received orthodontic treatment (ca 70% of American whites would benefit from orthodontics; Kelly and Harvey, 1977), some incisors were positioned ectopically or rotated, so film orientation perpendicular to the facial plane was not possible. These films were deleted, which accounts for variable sample sizes in the tables. The film with the best contrast was chosen for each of the four tooth types for each individual. Measurements were made with commercially available software (Adobe Photoshop 4.0; Adobe Systems, Mountain View, CA, USA) on a Macintosh platform. A periapical film was placed in a slide scanner (Polaroid SprintScan 35) and the image was digitized at 1200 dots/inch. After optimizing contrast and enlarging the computer image to approx. 8x, relevant variables were measured to the nearest 0.1 mm. (Extent of enlargement of the screen image had no effect on the actual measurements. Enlargement of the screen image was manipulated simply to aid in visualizing the landmarks.) Three characters were measured on each maxillary incisor: maximum mesiodistal crown width, maximum distal enamel thickness, and maximum mesial enamel thickness. Measurements were made perpendicular to the crown’s long axis.

Fig. 1. Schematic outlines of a maxillary lateral and central incisor showing measurements of maximum mesial and distal marginal enamel thicknesses and dentine thickness. Measurements were at the maximum crown width assessed perpendicular to the tooth’s long axis.

It can be argued that measurements from radiographs are less precise than if teeth were sectioned and the tissue types measured directly. Sectioning methods are not without their own technical problems, but, more importantly, the focus here was to provide normative data of clinical relevance and applicability. The availability of extracted sound permanent teeth for invasive analysis is minuscule compared to the opportunities to study patients’ teeth non-invasively.

Descriptive statistics for enamel thickness are listed in Table 1. None of the eight enamel thicknesses exhibited significant sexual dimorphism, so the data are presented for sexes pooled. All of these variables were normally distributed; that is, calculated from formulae in Sokal and Rohlf (1981), none of the tests for skewness or kurtosis approached statistical significance (=0.05). Enamel thickness was greater on the distal than the mesial side of each tooth, based on analysis of variance (randomized block design). The difference was highly significant (p<0.001) for both the lateral and the central incisors. Maximum marginal enamel thickness was about 0.1 mm greater on the distal sides of incisors.

The same statistical design was used to test between enamel thicknesses on the mesial aspects of the central and lateral incisors and, separately, between distal aspects of the two incisor types. Enamel was thicker on the distal margin of the central than the lateral incisor ( =1.05 mm vs 0.91 mm), which was a highly significant difference (p<0.0001). In contrast, mean thicknesses on the mesial aspects were almost identical, with a grand mean of 0.91 mm (p=0.92).

As there was no sex difference in marginal enamel thickness and it is well documented that crown diameters are appreciably larger in males than females, it follows that the sex difference rests with the dentine. Indeed, maximum crown widths and the widths of the dentine portion of the crown (obtained by subtraction of enamel thicknesses from crown width) were highly significantly larger in males. Sexual dimorphism averaged about 6.5% in these American whites, and it was slightly higher in the lateral than the central incisor and slightly higher for the width of the dentine than the width of the whole crown.

Because the enamel and dentine components of a tooth crown develop as a unified structure, there ought to be a correlation between them. Table 3 lists the correlation matrix between dentine and enamel thicknesses (mesial and distal widths combined here for brevity). A striking result is that those eight correlations involving enamel thickness and crown or dentine size were much higher in females than males. Just one of the eight correlations for males was significantly different from zero, whereas all eight of the female correlations were significant. Moreover, differences in the strength of the correlations achieved statistical significance in seven of these eight male–female comparisons. These female>male differences are of the same kind and at least as great as those noted by Garn et al. (1965) for crown size. In contrast, the last six comparisons in Table 3, which do not involve enamel thickness per se, are all moderate to high and are of similar strengths in the two sexes. Finally, marginal enamel thicknesses of the two incisor types exhibited a moderately high correlation (r=0.7), which is similar to correlations among maximum crown diameters.