Skeletal-versus soft-tissue-based cephalometric analyses: is the correlation reproducible?

OBJECTIVES: Dentofacial deformities can be analyzed by skeletal and soft tissue cephalometric analysis (CA). The aim was to evaluate the difference in reproducibility between both methods. MATERIALS AND METHODS: Lateral cephalograms of 112 patients (65 females and 47 males, 27.7 +/- 9.0 years) were oriented in natural head position (NHP) and digitized. The distances of skeletal (SNA, SNB, SnPog) and soft tissue (A', B' and Pog') landmarks relative to the respective norm values and the angles between the Nasion Sella line (NSL) and Frankfurt horizontal (FH) to NHP were measured for statistical evaluation and compared with respective data of an adult control group (CG) with class I occlusion and harmonic facial balance. RESULTS: The mean differences (mm +/- SD) of skeletal and soft tissue landmarks were -2.4 +/- 4.4 (A), -7.0 +/- 9.3 (B), -6.3 +/- 11.2 (Pog), -0.9 +/- 1.8 (A'), -4.7 +/- 6.2 (B'), and -6.1 +/- 7.8 (Pog'), respectively. Pearsons's correlation (r) between the measurements of SNA/A', SNB/B' and SNPog/Pog' were r = .158 (p = .092), r = .662 (p < .001) and r = .655 (p < .001), respectively. The mean (+/-SD) angles between NSL and FH to NHP were -9.8 degrees +/- 5 and 0.0 degrees +/- 3.9, respectively. CONCLUSION: Variability of cranial-based measurements could give a possible explanation for the high variation and the low reproducibility of skeletal cephalometric analysis with soft tissue measurements. Soft-tissue cephalometric analysis would probably improve facial analysis and treatment planning.