A cone-beam computed tomography study of hyoid bone position and airway volume in subjects with obstructive and nonobstructive adenotonsillar hypertrophy

The adenoids and palatine and lingual tonsils constitute the main component of the Waldeyer's ring, a lymphoid tissue complex around the pharynx that plays a crucial role in immunologic defense of the body. Moss' functional matrix theory states that the soft tissue directs skeletal tissue development. Consequently, nasal resistance due to adenoid or tonsillar hypertrophy is hypothesized¹  to cause developmental changes in the craniofacial complex.

The impact of mouth breathing on the dentoskeletal complex has been an area of debate and controversy for decades. Some scholars believe that mouth breathing is associated with alteration of craniofacial development and position of the hyoid bone,²  while others disagree.³  This is primarily because nasal airway inadequacy is subjective, and different authors judge breathing modes differently. Additionally, a recent systematic review and meta-analysis by Zhao et al.⁴  concluded that mouth breathing was linked to skeletal and dental developmental changes in children.

The association between mouth breathing and craniofacial structure deformities appeared in the orthodontic literature a long time ago. However, in mouth-breathing participants, the vast bulk of previous research asserted that the site of obstruction in the airway would not be a confounding factor, which, to some extent, might explain the controversial findings. However, there is still a lack of comprehensive knowledge on the particular impact of adenoid and tonsillar hypertrophy on the stomatognathic system. Additionally, several authors^5–7  linked the position of the hyoid bone to airway adequacy, but their assumptions were speculative and debatable. Hence, this study aimed to assess hyoid bone position and airway volume in subjects with obstructive and nonobstructive adenotonsillar hypertrophy and to investigate the relationship between airway volume and hyoid bone position in subjects aged 7 to 12 years.

This retrospective study was reviewed and approved by the Ethics Committee of Xi'an Jiaotong University. All participants' parents were informed, and signed written consent forms allowing use of their data for scientific purposes.

The sample consisted of 121 cone-beam computed tomographic (CBCT) images from Chinese children (65 females, 56 males) who were treated in the Orthodontic Department of Xi'an Jiaotong University from 2017 to 2022. Sample size was calculated using Minitab® v.18.1 software (Minitab Inc, State College, Pa), considering a power of 80%, a significance level of .05, a standard deviation of 4.2, and a 4.4-mm difference in hyoid bone vertical distance (HV) value.⁸  Twenty-two subjects in each group would be sufficient for the comparison. Preoperative CBCT scans of children aged 7 to 12 years with skeletal Class I (1 ≤ ANB° ≤ 4.9) or skeletal Class II (ANB° ≥ 5) and a normal vertical growth pattern (27 < FMA° < 37) were included. Subjects who displayed any of the following criteria were excluded: (1) previous history of adenoidectomy or tonsillectomy, (2) subjects who had undergone orthodontic treatment or had any syndromes related to the head and neck region, and (3) those with distorted CBCT scans with nonobvious landmarks.

CBCT images were taken in the natural head position. Participants were advised to remain in maximum intercuspation and to sit with a straight posture while maintaining the lips and tongue in the rest position, and they were instructed to breathe normally. All of the CBCT images were acquired with a cone-beam machine (i-Cat, Imaging Sciences International, Hatfield, Pa) using the following parameters: 120 kV, 5 mA, 14 ×17-cm field of view, 0.4-mm voxel size, and 8.9-second scan duration. The CBCT images were stored as Digital Imaging and Communications in Medicine (DICOM) files.

Dolphin Imaging Software (version 11.7, Dolphin Imaging & Management Solutions®, Chatsworth, Calif) was utilized to import the DICOM files. One investigator was responsible for all of the measurements, and this investigator was unaware of any of the participants' demographic data. To orient the CBCT images, the axial plane was adjusted to the Frankfort horizontal plane (FHP), the midsagittal plane was oriented in the midline of the patient passing through Nasion, and the coronal plane was oriented to pass through Porion, perpendicular to the axial plane.

All patients underwent clinical examination by a multidisciplinary team comprising an orthodontist and an otorhinolaryngologist. Initially, history was taken from the patients' parents to evaluate for the existence of bad sleeping habits: for instance, snoring, drooling on the pillow, and sleeping with an opened mouth. Subsequently, the Glatzel mirror test⁹  was performed. Otolaryngology examination was performed by a single investigator (LL) and included a physical examination followed by endoscopy, flexible nasopharyngoscopy, and nasopharyngeal x-ray.

The level of the adenoid or tonsillar hypertrophy was computed mathematically based on the CBCT images. The percentage of adenoid hypertrophy was calculated as AS/TNPA × 100; where AS indicates the linear distance extending from the posterior soft palate to the nearest point of adenoid tissues and TNPA refers to the total linear distance of the nasopharyngeal airway in the sagittal section. Percentage of tonsillar hypertrophy = TS/TOPA × 100, where TS is the narrowest distance between the tonsils in the coronal section and TOPA indicates the total linear distance of the oropharyngeal airway in the coronal plane (Figure 1). Subjects who displayed less than 25% of adenoid or tonsillar hypertrophy were considered nonobstructive, and those who displayed more than 50% of adenoid or tonsillar hypertrophy were considered obstructive. A similar approach was used by previous authors.^10–12  Subjects who met the inclusion criteria were divided into four groups, as follows: (1) control group (C-group) with nonobstructive adenoid or tonsillar hypertrophy, (2) adenoid hypertrophy group (AH-group), (3) adenoid and tonsillar hypertrophy group (ATH-group), and (4) tonsillar hypertrophy group (TH-group).

According to Nath et al.,¹³  the superior boundary of the airway was determined as a plane that ran posterior to the nasal spine (PNS) to the posterior wall of the pharynx (P plane), and the inferior boundary of the airway was determined to be a plane that ran parallel to the P plane and passed through the most superior point of the epiglottis. Subsequently, the airway was divided into an upper (retropalatal) segment and a lower (retroglossal) segment using a plane that was parallel to the P plane and passed through the uvula. Finally, the volumes were measured using Dolphin software, and the airway sensitivity was set to 73, as recommended by Alves et al.¹⁴  (Figure 2).

Hyoid bone linear measurements were HME, HEB, HC3, C3Me, HC3Me, HPNS, HH, and HV (Table 1; Figure 3).

All measurements for 30 randomly selected participants were repeated 2 weeks after the first measurement, and the intraclass correlation coefficient (ICC) was calculated to check the intrainvestigator reliability of the measurements. Dahlberg's formula¹⁵  was used to assess the measurement error.

The Shapiro-Wilk test was used to check the normal distribution of the data. Because the data showed nonnormal distribution, the nonparametric Kruskal-Wallis test was used for intergroup comparison, followed by pairwise comparison using the Mann-Whitney U-test with Bonferroni adjustment when a significant variable was detected.

Spearman correlation coefficient was performed to investigate the correlation between airway and hyoid bone parameters. Multiple linear regression analysis was used after the logarithmic transformation of the dependent variable values to build a model for the airway to identify the airway predictive parameters. SPSS software (version 25.0, SPSS Inc, Chicago, Ill) was used, and a P value < .05 was considered significant.

The ICC test revealed high intrainvestigator agreement, ranging between 0.93 and 0.99 for linear measurements and between 0.90 and 0.97 for volumetric measurements. The results of Dahlberg's formula were 0.81 to 1.60 mm and 165 to 323 mm³ for linear and volumetric measurements, respectively.

The baseline characteristics of the selected sample are shown in Table 2. No significant difference was found in the demographic variables or in the sagittal or vertical skeletal pattern variables (P > .05).

The intergroup comparison is displayed in Table 3. No significant difference was found between subjects with isolated adenoid or tonsillar hypertrophy in comparison to the control group with regard to hyoid bone position and airway parameters (P > .05). On the other hand, subjects in the ATH-group demonstrated a statistically significant decrease in HV, total airway volume (TA volume), and retroglossal airway volume (RG volume) values when compared to those in the C-group.

The ATH-group showed significantly lesser values of HV and all airway volumetric measurements when compared to the AH-group. While the comparison between the ATH-group and the TH-group showed significantly decreased TA volume and RG volume only, no significant difference in hyoid bone variables was observed.

Bivariate correlation analysis between airway and hyoid bone parameters is shown in Table 4. The correlation results in the C-group showed a moderate to strong positive correlation. Likewise, correlation results for the AH-group revealed a moderate to strong positive correlation. Correlation outcomes showed moderate to weak positive correlations for the ATH-group and TH-group.

Multiple linear regression analysis is shown in Table 5. The TA volume and retropalatal airway volume (RP volume) models were not as successful as the RG volume model. The RG volume model had a moderate potential for explaining the RG volume, representing 40% of the variance. The variables HV and age had a significant role in explaining the RG volume (beta = 0.4 and 0.17; P = .002 and 0.037, respectively). A linear regression equation was formulated as follows: Y = 2.099 + 0.009(X1) + 0.019(X2), where Y indicates the log of RG volume; X1 indicates the HV value; and X2 indicates age. The predicted and actual logs of RG volume are plotted in Figure 4.

In the current study, a collaborative approach was established between the Ear Nose and Throat and Orthodontic Departments to aid in the proper grouping of samples. Additionally, CBCT images were reported^10,12  to be a reliable and accurate tool for the assessment of adenoids and tonsils. Thus, CBCT images were used as an adjunctive tool for evaluating the extent of hypertrophy; images were only taken when they were expected to have an additive value for orthodontic diagnosis and treatment planning and following ALARA principles.¹⁶ 

The intergroup comparison revealed significantly different vertical hyoid bone position, described by the variable HV, between the C-group and ATH-group, indicating an elevated hyoid bone position in the ATH-group. This finding was similar to that of Chung et al.,²  who found that subjects with mouth breathing were characterized by an elevated hyoid bone. In addition, Chaves et al.⁷  reported that an elevated hyoid bone position was found in asthmatic patients. On the contrary, another investigation by Behlfelt et al.⁶  claimed that subjects with enlarged tonsils had a lower hyoid bone position. However, it is worth noting that these previous investigations used different methodologies than the current study, which, to some extent, might explain the controversial findings.

Jiang⁵  highlighted that a positive correlation between airway volumes and HME and HV was found, which was in agreement with the current correlation results. This could explain the previous findings, which reported¹⁷  airway and hyoid bone positional changes after orthodontic or orthognathic surgery treatment involving mandibular growth modification or advancement. Although the ATH-group showed the fewest correlations between airway and hyoid bone parameters, positive correlations were still found, especially in the RG volume (C3H: r = 0.447, HPNS: r = 0.355, HV: r = 0.486). These findings were similar to those of the previous case series performed by Riley et al.,¹⁸  who proposed the hyoid suspension technique as a surgical approach for treatment of obstructive sleep apnea. They concluded that suspending the hyoid bone antero-inferiorly contributed significantly in increasing the hypopharyngeal airway size, and a significant decrease in the apnea-hypopnea index (AHI) was observed postsurgically.

The muscular complex around the hyoid bone mainly consists of the suprahyoid and infrahyoid muscles. The suprahyoid muscles are largely responsible for lifting the hyolaryngeal complex and opening the upper esophageal sphincter during swallowing, while the infrahyoid muscles are responsible for maintaining the hyoid bone during swallowing to provide a stable foundation for the suprahyoid muscles to assist with mandibular depression. The middle pharyngeal constrictor muscle, which inserts into the hyoid bone, plays a vital role in maintaining airway patency. When the pharyngeal constrictor muscles are activated, the airway dilates at relatively low airway volumes, but it constricts at relatively high airway volumes.¹⁹  This suggests that alteration of hyoid bone position could affect muscle tonicity and, hence, the airway patency.

The relationship between obstructive sleep apnea and the hyoid bone has been investigated by several authors. Young and McDonald²⁰  found a significant correlation between all hyoid bone vertical measurements and the AHI. In addition, Chang and Shiao²¹  studied the craniofacial variables that could be contributing in obstructive sleep apnea; they highlighted that a positive correlation was found between the distance extending from the hyoid bone to the mandibular plane (H-MP) and AHI. A previous investigation by Pae and Harper²²  investigated factors that could be associated with a positive response to oral appliances for the treatment of obstructive sleep apnea; their findings showed that the hyoid bone moved superiorly after oral appliance therapy in subjects who demonstrated a greater decrease in AHI, suggesting that H-MP could be a useful marker for the favorable response to oral appliance treatment.

A collection of predictor variables that best explains variations in the dependent variable can be identified using linear multiple regression analysis. For TA volume, RP volume, and RG volume, every regression analysis was carried out independently. According to the results of the linear regression models, HV was a significant explanatory variable for the variation in RG volume. Another variable in the predicted model for RG volume was age. Previous research²⁰  underlined the significance of the vertical position of the hyoid bone in the severity of obstructive sleep apnea syndrome. Additionally, Schendel et al.²³  illustrated the potential impact of age on airway volume. Thus, it makes sense that the predictive model was able to adequately explain the variance in RG volume. The same variables utilized for TA volume and RP volume, in contrast to the RG volume model, provided low predictive value for the variances.

Despite the fact that the cross-sectional design of this study was not capable of assessing causation principles, a longitudinal design was not used because of the possible increased radiation associated with it. Another limitation of this study was that it did not include Class III subjects because of the lack of a large enough sample size. Variations of the hyoid bone position and, thus, the angulation and the relative muscular tone of the associated muscles in subjects with different skeletal phenotypes were previously reported.²⁴  Therefore, further investigations with more detailed grouping and a bigger sample size are recommended to better determine the effect of adenotonsillar hypertrophy on the craniofacial complex. In addition, the results of this study were limited to Chinese patients, and future studies should consider racial differences when evaluating the findings.

• A significant decrease in HV, TA volume, and RG volume was found in subjects with adenotonsillar hypertrophy compared to the control group.

• HV and age were significantly positively correlated predictors for explaining the RG volume.

This project was funded by (1) a general project from the field of Social Development in the Department of Science and Technology of Shaanxi Province, grant/award 2019SF-081; (2) the Science and Technology Project of Xi 'an, grant/award 20YXYJ0010(3); and (3) Clinical New Technology from Stomatological Hospital of Xi'an Jiaotong University in 2018.