A 2D geometric morphometrics analysis of mandibular morphological variability in class I, II and III

Objectives: Objectives: Evaluating mandibular morphological variability (MMV) in skeletal relations I, II and III using geometric morphometrics (2D-GM) and identifying mandibular anatomic features providing greater intra-group and inter-group variance.
Methods: Methods: This was a cross-sectional study; inclusion and exclusion factors gave eligibility criteria for obtaining data. Subjects gave informed consent. 2D-GM was used on lateral x-ray of 204 adults’ craniums (102 males, 102 females); 7 landmarks and 23 semi-landmarks were digitalised. Intra-observer error was measured by ANOVA-II after Generalized Procrustes analysis (GPA). Genders were independently grouped by Ward-cluster analysis on partial warps (PW). MMV was evaluated regarding the first four principal components (4PC) of each cluster. Multivariate multiple regression (MMR) analysis identified anatomic features supporting greatest intra- and inter-group variation.
Results: Results: Dimorphism was found (p<0.05) except for class III; 4 clusters were found in females and 3 in males. Class III concentrated in one cluster; classes I and II were mixed in remaining clusters. Differences were found for most clusters except between female C-D clusters. Male B-C clusters were more similar (H=2.62).
4PC gave variance ≥60% for both genders. Intergroup tendency was observed regarding MMV direction and magnitude in both genders for each cluster. Third PC gave greater explanatory power (t=-3.4;p=0.0007).
Conclusions: Conclusions: Obtained clusters matched prior craniofacial classification by midface, skull base and jaw variables where class III concentrated in one cluster while I and II mixed in the remainders. Variation direction and magnitude for mandibular anatomical features within clusters suggested a convergent response via MMV, probably determined by selective pressure on function. GM allows identifying and visualising morphological variations; this could form the basis for assessing real anatomical effects which could explain such craniofacial patterns.