Three-dimensional phenotype characteristics of skeletal class III malocclusion in adult Chinese: a principal component analysis-based cluster analysis.

Abstract

Skeletal class III malocclusion has a diverse and complicated aetiology involving environmental and genetic factors. It is critical to correctly classify and define this malocclusion to be diagnosed and treated on a clinically sound basis. Thus, this study aimed to provide reliable and detailed measurements in a large ethnically homogeneous sample of Chinese adults to generate an adequate phenotypic clustering model to identify and describe the skeletal variation present in skeletal class III malocclusion. This is a retrospective cross-sectional study in which 500 pre-treatments cone-beam computed tomography (CBCT) scans of patients with skeletal class III malocclusion (250 males and 250 females) were selected following specific selection criteria. Seventy-six linear, angular, and ratios measurements were three-dimensionally analysed using InVivo 6.0.3 software. These measurements were categorised into 47 skeletal, 18 dentoalveolar, and 11 soft tissue variables. Multivariate reduction methods: principal component analyses and cluster analyses were used to present the most common phenotypic groupings of skeletal class III malocclusion in Han ethnic group of Chinese adults. The principal component analysis revealed eight principal components accounted for 72.9% of the overall variation of the data produced from the seventy-six variables. The first four principal components accounted for 53.37% of the total variations. They explained the most variation in data and consisted mainly of anteroposterior and vertical skeletal relationships. The cluster analysis identified four phenotypes of skeletal class III malocclusion: C1, 34%; C2, 11.4%; C3, 26.4%; and C4, 28.2%. Based on three-dimensional analyses, four skeletal class III malocclusion distinct phenotypic variations were defined in a large sample of the adult Chinese population, showing the occurrence of phenotypic variation between identified clusters in the same ethnic group. These findings might serve as a foundation for accurate diagnosis and treatment planning of each cluster and future genetic studies to determine the causative gene(s) of each cluster.