Characterization of third molar morphometric variables

The third molar is a tooth of anatomical, surgical, prosthetic and forensic dental interest. However, there is currently a lack of updated data regarding its morphology. The aim of this study was to determine the morphometric features of third molars and their predictive capability as regards dental arch and side. Two calibrated operators (ƙ = 0.83) determined the cervicalocclusalvestibular (COV), cervicalocclusalpalatal (COP) and occlusalapical (OA) distances, mesiodistal (MD), and vestibularpalatal (VP) diameters, number of roots (R) and number of cusps (C) of 961 cadaveric third molars, both upper (n = 462) and lower (n = 499), using a CONCOR 050 thin mandible caliper (resolution 0.01 mm). Median and range for each variable were calculated and compared using Mann Whitney nonparametric test (p < 0.05). Multivariate cluster analysis was used to determine the predictive capability of each variable for dental arch and side. For upper molars (UM), 50.6% were from the right side (RS) and 49.4% from the left side (LS), while for lower molars (LM), 60.9% were from the RS and 39.1% from the LS. No significant difference was found in the study variables in LM according to side. For UM, MD diameter (10.90 mm), COP(7.42 mm) distance and number of R (3) were significantly higher (p < 0.05) forRS, and number of C (3) was higher (p < 0.0001) for LS. They were also significant predictive grouping factors for side. For dental arch, OA (17.84 mm) and COV (7.60 mm) distances, MD (11.26 mm) diameter and the number of C (5) were significantly higher (p < 0.0001) for LM, while VP (10.84 mm) and COP (7.34 mm) distances, and the number of R (3) were significantly higher (p < 0.0001) for UM. These variables were significant predictive factors for dental arch. Despite the morphometric heterogeneity of third molars, there are intrinsic parameters with predictive capability for dental arch and side, but it would be advisable to supplement this study with data from topographic occlusal variables in order to validate their predictive capability.