We enrolled 15 DS subjects (mean age, 9.43 ± 0.38 years; 8 boys, 7 girls) and 15 non-DS subjects (mean age, 9.51 ± 0.40 years; 8 boys, 7 girls) who visited the Department of Orthodontics, Showa University Dental Hospital, from August 2016 to August 2020. Subjects with a history of surgery-related cleft lip and palate, tracheomalacia, laryngomalacia, tonsillectomy, adenoidectomy, or previous orthodontic treatment were excluded. The subjects were classified according to their anteroposterior skeletal pattern using the ANB angle: Class I, -1° ≤ ANB < 4°; Class II, ANB ≥ 4°; and Class III, ANB < -1°40,41. Vertical skeletal patterns were classified using the angle of the Frankfort horizontal plane (FH) to the mandibular plane (MP) (FH/MP): hypodivergent, FH/MP < 22°; normodivergent 22° ≤ FH/MP < 30°; and hyperdivergent FH/MP ≥ 30°41. Patient demographics are shown in Tables 4 − 1, 4 − 2, 4 − 3 and Fig. 1. Written informed consent was obtained from all subjects and/or their parents and the study was approved by the Ethics Committee of Showa University Dental Hospital, and all parts of the study were performed in accordance with the guidelines of the Declaration of Helsinki (approval number: SUDH0077).
CBCT examination was carried out for all subjects to evaluate skeletal malocclusion, impacted permanent teeth, and the skeletal relationship between the maxilla and mandible. CBCT images were taken with a KaVo 3DeXam (KaVo Dental, Biberach, Germany). The scanning conditions were set at 120 kV and 5 mA, the voxel size was 0.4 mm, and the scanning time was 8.9 seconds. Subjects were seated comfortably, with a natural head position after adjustment of the chin rest, and were asked to bite but not move or swallow during the scan. The CBCT images were saved in Digital Imaging and Communication in Medicine format.
To evaluate the nasopharyngeal structures, the CBCT images were examined using Invivo5 dental radiology software (Anatomage, San Jose, CA, USA). The images were reoriented by using the FH plane as a reference plane to standardize the measurements and to minimize errors (Fig. 2)19,20. The FH plane was defined by the right and left porions (the most laterosuperior point of the external auditory meatus) and the right and left orbitales (the most inferior point of the lower margin of the bony orbit). Five cross-sectional planes: the anterior nasal plane (Ana), posterior nasal plane (Pna), upper pharyngeal plane (Uph), middle pharyngeal airway (Mph), inferior pharyngeal airway (Iph), and five volumes of the pharyngeal airway were configured (Table 5). To separate and extract the airway space from the craniofacial region, we used a threshold tool with the histogram adjusted as a guide to -523.9 Hounsfield units20,42. We calculated the defined nasopharyngeal airway volumes in cubic millimeters, and the airway width, length, and area in the sectional views (frontal and axial) in the cross-sectional lines (Figs. 2 and 3).
Statistical analysis
Statistical analyses were performed using SPSS Statistics, version 25 (IBM Corp., Armonk, NY, USA). To assess intraoperator error, all measurements on the CBCT images were remeasured at a separate session with a 2-week or more interval under identical conditions. The measurement error was estimated according to ICC assessment. The Shapiro–Wilk test and Leven’s test were used to confirm normality and equality of variance among the measurements in this study. ANCOVA with Bonferroni post hoc pairwise comparison tests for the corrected means were used to compare the 26 measurements of the volume, area, and length of the nasopharyngeal airway among the groups and between individual differences, with body height and body weight as covariates. Moreover, as covariates with the ANB angle and the mandibular plane angle, ANCOVA with Bonferroni post hoc pairwise comparison tests were also conducted. Significance probability adjusted with Bonferroni was set at p < 0.019.
Power analyses were performed using G*Power Ver. 3.1.9.4 (Franz Faul, Universität Kiel, Germany) to calculate prior required sample sizes at a two-sided significance level of 5% and a power of 80%. A sample size of at least 14 patients for each group was calculated to detect the difference in total airway volume as a primary measurement between control and DS patients, assuming mean volumes of 21,000 mm3 and 17,000 mm3, respectively, with a common standard deviation of 3,600 mm3. Post hoc power analysis was also carried out using G*power version 3.1.9.6 (Franz Faul) for calculation of detection power at a two-sided significance level of 5% and a sample size of 15 in each group.