Are teeth useful in estimating stature?

Estimating stature is an important step in reconstructive identification of skeletonized and dismembered human remains. While numerous body parts such as the skull and long bones have been used for the purpose, the dentition has seldom been applied. The present study has ventured to ascertain the usefulness of tooth crown measurements in stature prediction. Buccolingual and mesiodistal dimensions of all teeth (except third molars) and stature measurements were obtained from 95 living adults (47 females, 48 males). Correlation analysis revealed that 21 of the 56 tooth crown variables had a low albeit statistically significant correlation to stature (p < 0.05); correlation matrix computed for the crown variables showed significant inter-correlations between most teeth (problem of multi-collinearity). Therefore, instead of regular least square regression analysis, ridge regression was performed for the dentition, which revealed a moderate but statistically significant correlation to stature (R = 0.68; p < 0.0001). The ridge regression equation derived had a standard error of estimate (SEE) of 8.09 cm. The multiple correlation for tooth dimensions is lower to, and the SEE larger than, most other body parts. The moderate correlation is probably due to early completion of growth of tooth crowns vis-à-vis other parameters such as long bones that mature later and have a higher stature-correlation. This indicates that the dentition may be used only as a supplement to more robust indicators of stature.