Do bivariate and multivariate cephalometric analyses lead to different results concerning the skeletal cause of postnormal occlusion

In most cephalometric studies on the craniofacial pattern of untreated Class-ll-malocclusion subjects mean values were bivariately compared with the mean values of subjects with normal occlusion or Class I malocclusion, or with standard values. A multivariate approach, which does not merely compare two variables at a time but which analyses several skeletal variables at once, is lacking. The objective of the present study was in a first step to bivariately compare the sagittal and vertical skeletal pattern of Norwegian untreated postnormal occlusion patients with Class I children and in a second step to compare the results of the bivariate analysis with the results of a multivariate analysis on the same subjects. Previous studies showed in the majority of cases a retropositioned mandible rotated to the posterior with an indefinite positioned maxilla. A bivariate comparison of 138 Norwegian subjects with Class II/1 malocclusion and 80 children with Class I occlusion from the same region (mean age of both 9 years) confirmed the posterior position of the mandibles. However, a multivariate analysis (harmony-box-concept) of the same subjects yielded a different result. Simultaneously taking the sagittal and vertical position of both jaws and the cranial base flexure angle (NSBA) into consideration, the comparison of the two groups demonstrated a forward positioned maxilla as the skeletal cause of the Class II/1 malocclusion. The mandibles were on average normal positioned instead of retropositioned relative to other skeletal structures of the viscerocranium. Apparently standard bivariate tests and multivariate analysis methods may lead to different results which also have to be considered by the functionally treating orthodontist.