Changes in hyoid bone position following rapid maxillary expansion in adolescents

Abstract

Objective:

To examine changes in hyoid to mandibular plane distance (H-MP) and tongue length (TL) between children who had orthodontic treatment with and without rapid maxillary expansion (RME).

Materials and Methods:

Lateral and frontal cephalograms of 138 patients treated with RME and 148 controls treated without RME were used to measure pretreatment (T₁) and posttreatment (T₂) intermolar (IM) distance, lateronasal width (LNW), H-MP, and TL. Medical histories were used to collect demographic information, history of mouth breathing, difficulty breathing through the nose, and previous adenotonsillectomy. Groups were group-matched for age and gender. Descriptive statistics were calculated. Group means were compared using t-tests and chi-square statistics. Reliability was estimated using intraclass correlations and kappa statistics. Statistical significance was set at P < .05.

Results:

At T₁, the RME group showed smaller LNW (24.83 ± 1.99 vs 26.18 ± 2.05) and IM (50.17 ± 2.3 vs 51.58 ± 2.83). The distance from H-MP was longer in the RME group (15.69 ± 3.95 vs 13.86 ± 3.4). Mean changes (T₂ − T₁) in the RME group were increased LNW (+2.48 ± 1.38 vs +0.94 ± 1.11 for the non-RME group) and IM (+3.21 ± 1.72 vs +0.98 ± 1.67). The mean change (T₂ − T₁) in H-MP for the RME group was −0.68 ± 3.67 compared with +1.1 ± 2.96 for the non-RME group. Mean changes for TL were not statistically significant. No significant differences were noted at T₂ between groups for LNW, H-MP, or TL.

Conclusions:

In this sample, RME produced significant changes in H-MP, and TL was unaffected.

INTRODUCTION

Orthodontic treatment provides many benefits to patients. Improved esthetics and masticatory function are the best known. Recent evidence suggests that devices designed to protrude the mandible may be effective for the treatment of breathing problems and snoring.1 The goal of such therapy is to modify the position of upper airway structures in efforts to enlarge the airway and reduce its collapsibility.

Another way to reduce upper airway collapsibility is to reduce the resistance to airflow within the airway passage. Maxillary transverse deficiency reduces the cross-sectional area of the airway, leading to increased nasal resistance. Previous studies have demonstrated a reduction in nasal resistance following rapid maxillary expansion (RME) with banded or bonded appliances.2–4 Evidence suggests that RME may reduce the apnea hypopnea index in children with sleep apnea.5–7 Improved nasal airflow and resolution of obstructive sleep-disordered breathing have been reported in adults undergoing surgically assisted RME.8 RME treatment improves nasal breathing by significantly increasing total minimum cross-sectional area and total nasal volume.9

Evidence indicates that hyoid bone position may be affected by upper airway resistance. Verin et al.10 found that transpalatal resistance was correlated with greater hyoid to mandibular plane distance. It has also been shown that the hyoid bone becomes progressively lower as airway resistance increases.11 Hyoid bone position changes with age. According to Tourne,12 the hyoid bone descends during growth and maintains its position between C3 and C4. Taylor13 also demonstrated a steady descent of the hyoid bone during adolescent growth. Nelson et al.14 found that although snoring subjects had lower hyoid bone position at all ages, the hyoid position became lower with increasing age, regardless of snoring status.

Because increased nasal resistance is associated with maxillary transverse deficiency, and because RME decreases nasal resistance, it is possible that RME affects hyoid bone position. In addition, increased tongue length (TL) has been associated with obstructive sleep apnea,15 and RME has been recommended for treatment of obstructive sleep apnea syndrome (OSAS). The purpose of this study was to examine differences in hyoid to mandibular plane distance (H-MP) and TL between children who had orthodontic treatment with and without RME.

MATERIALS AND METHODS

The treatment records of 630 children, who were active patients between 2001 and 2005 at an established orthodontic graduate clinic in the Midwest, were examined. Patients in the surgical, adult, and craniofacial anomalies specialty clinics were excluded. Of the remaining 556 charts, 156 patients whose treatment plan included RME (Hyrax, Haas, or bonded expanders activated at least every other day) followed by fixed appliances, made up the experimental group. The control group (165 patients), consisting of patients treated with braces alone, was randomly chosen from the remaining charts and was group-matched by age and gender.

Inclusion criteria for both groups were as follows: (1) patients younger than age 18 at pretreatment, (2) patients who completed medical history forms, and (3) patients with complete pretreatment and posttreatment radiographic records. Patients who had not completed their treatment, who underwent tonsillectomy/adenoidectomy or any other surgical procedures during the course of the orthodontic treatment, or who appeared to have atypical posture on the lateral radiograph (eg, teeth not together, swallowing as evidenced by the morphology of the soft palate on the radiograph) were excluded. On the basis of these criteria, 53 patients were excluded.

Confirmation of group-matching by age and gender was supported by the absence of statistical differences in these assessments. The sample treated with RME consisted of 138 individuals (55 boys and 83 girls). Average age at the start of treatment (T₁) was 13 years 2 months ± 1 year 6 months. Average age at the end of treatment (T₂) was 15 years 8 months ± 1 year 6 months. The sample treated without RME consisted of 148 children (53 boys and 95 girls). Average age at T₁ was 13 years 6 months ± 1 year 6 months. Average age at T₂ was 15 years 10 months ± 1 year 7 months. The duration of orthodontic treatment in both groups was approximately 2 years 5 months.

The patient medical history and the treatment record were used to collect the following information:

• Patient's age at the time of initial and final records

• Gender

• Race

• Mouth breathing at T₁

• Difficulty breathing through the nose

• Adenoids/Tonsils removed before treatment.

All lateral and frontal cephalograms at T₁ and T₂ were adjusted for magnification. Average magnification for lateral and frontal radiographs was 14% and 8%, respectively. Magnification was calculated by measuring the actual and perceived dimensions of the nasion piece and the ear piece. Tracings of initial and final frontal and lateral radiographs were used to collect the following cephalometric measurements with a digital caliper:

• Pretreatment and posttreatment shortest perpendicular distance from the most anterior-superior point of the hyoid bone to the mandibular plane (Figure 1)

• Pretreatment and posttreatment TL measured from the base of the epiglottis to the tip of the tongue behind the lower incisors (Figure 1)

• Pretreatment and posttreatment intermolar (IM) width (the most prominent lateral point on the buccal surface of the upper first molar) measured from the frontal cephalogram (Figure 2)

• Pretreatment and posttreatment lateronasal width (LNW) measured as the distance between the most lateral points of the nasal cavity (Figure 3).

Figure 1.

Tongue length measured from the base of the epiglottis to the tip of the tongue behind the lower incisors. Hyoid distance measured from the most anterior-superior point of the hyoid to the mandibular plane.

Figure 2.

Intermolar distance measured between the most prominent lateral surfaces of the maxillary first molars.

Figure 3.

Lateronasal width measured from the most lateral points of the nasal cavity.

All measurements were completed by a single operator. Ten percent of the sample was randomly identified, and cephalometric measurements were repeated at a 2-week interval. Categorical data were reentered for the same 10% of the sample. Paired t-tests used to compare continuous variables revealed no statistical difference (P < .05). Intraclass correlation coefficients for all continuous variables were above 0.91. The kappa statistic was used to compare agreement of the categorical variables. All categorical variables had perfect agreement (coefficient of 1), except race, which had a coefficient of 0.71.

Statistical Analysis

Histograms of continuous variables showed that they were normally distributed. Descriptive statistics were obtained for cephalometric measurements at T₁ and T₂. Independent sample t-tests for comparisons of means between groups were used to evaluate the changes, and paired t-tests were used to examine changes within groups. Chi-square statistics was performed for categorical variables. Computer software (Statistical Package for the Social Sciences [SPSS], version 10, SPSS Inc., Chicago, Ill) was used to complete statistical computations.

RESULTS

Age, gender, and race distributions for the two groups are given in Table 1. At T₁, statistical differences were observed between RME and non-RME groups (Table 2). The RME group showed a deficiency in LNW and in IM distance when compared with controls. The distance from H-MP was also longer in the RME group. Unlike the cephalometric variables, none of the categorical variables (tonsils and adenoids removed, mouth breathing, and difficulty breathing through nose) showed significant differences at T₁ (Table 3).

Table 1.

Demographic Data

Table 2.

Comparison of Expansion vs Nonexpansion Groups at T₁

Table 3.

Categorical Variables at T₁

RME induced significant changes in the transverse dimension (Tables 4 and 5). The RME group similarly demonstrated (T₂ − T₁) a greater increase in LNW and in IM width. The H-MP distance decreased in the RME group and increased in the non-RME group.

Table 4.

Comparison of Groups (T₂ − T₁)

Table 5.

Comparison Within Groups (T₂ − T₁)

At T₂, no statistical difference was noted in LNW between the groups (Table 6), and IM was greater in the RME group. The between-group difference in hyoid bone position observed at T₁ was indistinguishable at T₂. TL showed no differences between groups at T₁ or T₂.

Table 6.

RPE vs Non-RPE Groups at T₂

DISCUSSION

This retrospective cohort study evaluated differences in changes in hyoid bone position and TL among adolescents undergoing orthodontic treatment with and without RME. Both groups were matched for age and gender. A statistically significant difference in time period between initial and final records was noted between groups, with the RME treatment taking about 2 months longer.

We did not find a difference in mouth breathing between the two groups. Previous studies3,4 showed that subjects receiving RME for orthodontic purposes had a much higher nasal resistance before treatment than subjects not receiving RME. In those studies, the median nasal resistance of the RME group before treatment was well above the value (5.5 H₂O/L/s) at which clinical mouth breathing was observed, and subjects receiving orthodontic treatment without RME had lower nasal resistance well below this value. Comparison16 of mouth-breathing and nasal-breathing children revealed that a narrow upper arch, measured as IM distance, had a significant correlation with mouth breathing. In our study, neither mouth breathing nor measures of nasal resistance were evaluated clinically. Instead, they were recorded from the medical history form filled out by parents. The most likely explanation for our findings is that parents were unaware of the child's difficulty breathing through the nose or did not deem it important to report them as mouth breathers.

No statistical difference was seen between groups in having adenoids and tonsils removed before treatment. The overall number of patients reporting these procedures in our sample was too small to allow any conclusions. However, it should be noted that enlarged tonsils are associated with lower hyoid bone position,17 and tonsillectomy leads to reduced nasal resistance.18 Because of the possible confounding effects of adenotonsillectomy on the airway, patients who had these procedures performed during orthodontic treatment were excluded.

Treatment with RME produced 3.21 ± 1.72 mm of IM width increase and 2.48 ± 1.38 mm of LNW increase. These changes were significantly greater than in the non-RME group. A significant (P < .01) correlation of 0.464 between IM increase and nasal width increase was noted. Because the maxilla expands in the shape of a V with the apex at the nasal floor, it is logical that IM expansion is greater than nasal expansion. Although changes in molar and nasal width are consistent with previous reports, differences observed at T₁ suggest that a susceptibility bias may have influenced our results.

A significant difference in hyoid bone position was seen between the groups at T₁, with the hyoid bone being nearly 2 mm lower in the RME group. During the course of treatment, the hyoid to mandibular plane distance increased in the non-RME group, as might be expected, because the hyoid normally descends during maturation.13,14 In contrast, RME treatment resulted in a decrease in this distance, to the point that no differences in hyoid position were distinguishable at T₂. Although the differences at T₁ might suggest a susceptibility bias, it is highly unlikely that clinicians treating these cases used H-MP as a diagnostic criterion. Thus, it is more likely that the effect on H-MP observed in the RME group was the result of the treatment. Future studies with long-term follow-up will be required to allow conclusive statements on the stability of the changes noted.

Airway geometry and lumen suction pressure can lead to obstruction in the area of the oropharynx. The narrower the airway, the more susceptible it is to collapse.19 Several anatomic features contribute to the narrowing: transversely deficient maxillae, enlarged adenoids and tonsils, and enlarged tongues in obese individuals. Increased lumen suction pressure is created when the forces generated by inspiratory muscles produce greater positive transmural pressure favoring upper airway collapse. Increased resistance in the nasal airway, for example, will create greater suction pressure in the oropharynx.

The relationship between oropharyngeal obstruction and hyoid bone position is evident from previous studies that reported an association between low hyoid bone, long tongues, and obstructive apnea, as well as snoring in adults.14,15,20 Similar findings have been reported for children with sleep apnea21 and snoring.22 If airway obstruction is related to narrow maxillae, this suggests that reducing maxillary transverse deficiency could improve airway function and potentially normalize hyoid bone position. This could be of significant importance to the craniofacial practitioner, as the hyoid bone is an anatomic structure that is easily identified and evaluated on lateral cephalograms.

Tongue length, which may also affect airway size and lead to obstruction of the oropharynx,19 showed no differences between groups at T₁ or T₂. Previous studies have shown that greater TL is associated with a higher Respiratory Disturbance Index.15 In our study, TL acts as a covariate in that it can affect the hyoid position but might not be related to the narrow maxilla. This may serve as a possible explanation for the lack of statistical difference in TL at T₁. It is unlikely that RME can induce changes in TL. TL increased slightly in both groups; however, this finding was likely due to normal growth. Cohen and Vig23 documented the increase in tongue mass in a longitudinal study of 50 individuals. They found that the growth rate slows down as the individual gets closer to 18 years of age. Our results suggest that RME does not influence TL.

Limitations

A limitation of our study was that patients in the RME group were in the group because the clinicians treating them made a diagnosis of transverse maxillary deficiency and used an RME appliance for orthodontic expansion. We could not control the decision made, nor do we have knowledge of formal guidelines followed. However, IM width and LNW were both significantly smaller in the RME group, which is consistent with a diagnosis of transverse maxillary deficiency.

As was mentioned previously, another limitation was a susceptibility bias. Our control group consisted of orthodontic patients receiving treatment not involving RME. To better isolate the effects of RME, patients with maxillary transverse deficiency not receiving any treatment could have been used as controls. However, in the clinic setting, records are taken only on patients committed to treatment. It would also be hard to justify delaying expansion because the efficiency of RME decreases with age. Using orthodontic patients as controls presents an advantage as it helps to eliminate the effects that orthodontic treatment (other than RME) may produce. Final records were taken at the end of orthodontic treatment in our study. Perhaps to better isolate the effects of RME on hyoid bone position and tongue length, records immediately after expansion was completed could be useful. However, using the final records has the advantage of evaluating changes in hyoid bone position 2 years after expansion.

CONCLUSIONS

• Before treatment, the hyoid to mandibular plane distance is greater in patients with narrow maxillae requiring RME.

• Following treatment, the hyoid bone to mandibular plane distance increased in non-RME subjects and decreased in subjects treated with RME.

• RME treatment tends to normalize hyoid bone position.

• No evidence suggested that TL was related to RME and/or transverse maxillary deficiency.