Craniofacial shape from pre- to post-adolescence

Craniofacial growth demonstrates significant variation and is difficult to predict. The aim of the present investigation was twofold: (1) to assess the association (covariation) between craniofacial shape at pre- and post-adolescence and (2) to evaluate if pre-adolescent craniofacial shape is related (covaries) with growth magnitude and direction.

One hundred fifty subjects (86 males and 64 females) untreated orthodontically were selected from AAOF Craniofacial Growth Legacy Collection. Each subject had cephalograms taken before 9 (pre-adolescent stage) and after 15 years of age (post-adolescent). Fourteen curves comprising 123 points (10 fixed and 113 sliding semilandmarks) comprehensively covering the craniofacial skeleton were digitally traced on each cephalogram. Procrustes alignment, principal component analysis, 2-block partial least squares (2B-PLS) analysis, and regression analysis were done after sliding the semilandmarks to minimize bending energy.

The first 16 principal components (PCs) were non-trivial and explained 85.2% of total shape variability in the sample. PC1 depicted mainly variability in the vertical direction, PC2 represented mostly variability in the saddle angle and in the antero-posterior position of the mandible, and PC3 depicted primarily variability of the mandibular shape (steep versus flat mandibular plane). The covariation between pre- and post-adolescent facial shape was statistically significant, both in the pooled sample (RV coefficient = 0.604) and in boys (RV = 0.639) and girls (RV = 0.629). The pre-adolescent shape was weakly associated with the magnitude of facial change-2-block PLS analysis demonstrated that blocks 1 and 2 were independent (P = 0.118, RV = 0.035).

The pre-adolescent shape of the craniofacial complex explained approximately 60% of the post-adolescent shape of the craniofacial complex; however, the relationship between pre-adolescent shape of the craniofacial complex and magnitude of its change was weak.