A 3D geometric morphometric analysis of craniofacial sexual dimorphism in the contemporary Colombian population

Objectives: Quantitatively describing craniofacial dimorphic variables and their allometric influence.

Methods: Morphogeometry (MG) was used on 3D images of dead Colombian adults’ craniofacial bones; 158 landmarks were digitised. A generalized Procrustes analysis (GPA) was made, the measurement method being validated by two-way ANOVA classification without the interaction of repeated-measures and intraclass correlation coefficients. Permutational multivariate analysis (for discriminant functions, canonical analysis and regressions multivariate) was used for evaluating the significance of the results.


Results: Distance in space (coordinates) represented dissimilarities in principal component analysis (CPA), thereby facilitating group exploration and analysis and revealing evident overlapping between males and females. The differences were plotted as anatomical deformations describing patterns. Canonical correlation analysis of the first 6 CPA revealed the sample’s natural grouping tendency. Discriminant analysis involving crossed validation determined an acceptable percentage of correct assignation regarding unnamed individuals’ shape and form. Multivariate regression between centroid configuration and size gave low correlation levels. The average effect of size was evaluated for determining the degree of SD, males being effectively larger than females (5.17% SD index).

Conclusions: Gender differences regarding contemporary Colombian adults’ craniofacial skeletons were significant concerning size, formation and form, having a slight but significant allometric component