Nasal cavity structural anomalies in children and adolescents at high risk of sleep-disordered breathing: An exploratory cone-beam computed tomography study


In this study, we investigated the presence of structural anomalies in the nasal cavity (deviated nasal septum [DNS] and turbinate hypertrophy [TH]) in patients at high risk or not of sleep-disordered breathing (SDB).


A retrospective study considering available cone-beam computed tomography scans of 99 patients was conducted. Dolphin Imaging software (Dolphin Imaging and Management Solutions, Chatsworth, Calif) was used to process the craniofacial scans. A pediatric sleep questionnaire (PSQ) was used to suggest a high risk of SDB. Subjective and objective assessments of DNS and TH were considered.


Good to excellent intrareliability and interreliability were attained. The prevalence of a PSQ score suggestive of a high risk of SDB in this sample was 59%. The prevalence of subjective DNS and TH assessment was 64% and 70%, respectively. In contrast, on the basis of objective assessments, 27% of patients presented with DNS and 25% with TH. Cross-tabulation of DNS and TH with PSQ score indicated a statistically significant association between subjective DNS and subjective TH and subjective TH and positive PSQ. A positive correlation between age and subjective and objective DNS assessments was also observed.


Older patients are more likely to present with DNS. Only the presence of subjectively determined TH in patients is associated with a high risk for SDB. The study reveals that assessment of DNS and TH using cone-beam computed tomography imaging is not likely suitable to strongly suggest patients at high risk for SDB. DNS subjective assessments were capable of identifying less than 5% of deviation.


  • Deviated nasal septum and turbinate hypertrophy were assessed in children using CBCT.

  • Findings from CBCT scans were correlated with risk for obstructive sleep apnea.

  • Neither septum deviation nor turbinate hypertrophy correlated with high risk for obstructive sleep apnea.

  • CBCT can be used in conjunction with a medical history and clinical assessment.

The nasal cavity is divided into 2 sides by the nasal septum. The nasal septum, a cartilaginous tissue, provides structural support for the upper nasal cavity. Abnormalities to the septum could significantly affect the nasomaxillary complex growth and breathing in humans. One such abnormality is a deviated nasal septum (DNS), sometimes associated with nasal airway obstruction. A DNS is defined by an anterior cartilage deformity of the quadrilateral septal cartilage or by a combined septal deformity that involves all septal components. The septum deviation varies from a C- or S-shaped septum to an irregular, angulated, dislocated septum. The consequence of DNS leading to nasal airway obstruction include snoring and increased nasal resistance.

The nasal cavity also contains 3 pairs of turbinates (superior, middle, and inferior) thought to contribute to the regulation of nasal airflow and humidification. Turbinate hypertrophy (TH) (enlargement of turbinates) is also associated with nasal airway obstruction.

DNS and TH are both considered potential contributors to obstructive sleep apnea (OSA), the most common form of sleep-disordered breathing (SDB), although their relative contribution is still being explored. SDB is a disorder that affects around 1%-5% of children. , If untreated, persistence of SDB could be associated with life-threatening cardiovascular, metabolic, and neurocognitive complications in addition to a considerate economic health burden. Studies investigated adult patients (mean age, 40 years) with DNS and suggested that DNS leading to severe nasal obstruction could lead to breathing abnormalities during sleep. In addition, research also suggests that to correct for DNS, a septoplasty may be indicated to alleviate patients from snoring. , A recent meta-analysis indicated that adult patients who underwent isolated nasal surgery for OSA improved an average of −4.15 (95% confidence interval [CI], −6.48 to −1.82) in the apnea–hypopnea index scores. However, whether such a surgical approach could help patients with DNS that have developed OSA is unclear. In addition, several studies investigated nasal airway abnormalities as predictors of pediatric OSA, although the extent DNS or TH contribute to OSA in children or adolescents is still unclear. Should such abnormalities significantly contribute to the risk for OSA in children or adolescents, it would be crucial to better understand the onset and presentation of these abnormalities in children and adolescents.

A validated pediatric sleep questionnaire (PSQ) was used as a prescreen assessment to identify children or adolescents at high risk for SDB. This screening approach was used before a full sleep assessment to better identify children or adolescents that are more likely to be affected and alleviate the current significant wait times for full diagnostic assessment. The PSQ is not specific in determining the etiology of SDB. Thus, we investigated the use of already available cone-beam computed tomography (CBCT) scans as an additional assessment tool to identify potential anatomic deviations that may contribute to facilitating an increase in risk for SDB among children. If there were a strong link, then a multidisciplinary evaluation and management of the structural abnormalities in patients with SDB would lead to a better overall treatment planning and outcome and possibly prevent secondary morphologic and developmental complications.

On the basis of the stated discussion, this study aimed to assess the potential association between structural anomalies in the nasal cavity (DNS and TH) with at high risk for SDB in a convenient sample of children and adolescents aged 6-17 years. In addition, this study aimed to identify the assessment capability of the subjective and objective methodology using CBCT scans to assess DNS and TH.

Material and methods

A retrospective cross-sectional clinical study was performed using already available CBCT scans of children and adolescents. This cohort was identical to the PSQ cohort used in Eimar et al, with the exception of 2 participants for whom CBCT scan quality was deemed insufficient for assessment. Ethical approval was obtained from the University of Alberta, Edmonton, Alberta, Canada (no. Pro00081755). Informed consent was already available from patients or guardians that their scans and information may be used for research purposes. The CBCT scans of 99 children or adolescents who visited the Graduate Orthodontic Program Clinic at the University of Alberta were evaluated. CBCT scans were obtained on an iCAT (Imaging Sciences International, Hatfield, Pa) scanner with 0.3-mm voxel size at 120 kVp, 18.54 mA, and exposure time of 8.5 seconds. The field of view was 16 × 6-cm, resulting in an effective radiation dose of approximately 35 microsieverts). Images were reconstructed and stored using Dolphin 3D software (Dolphin Imaging and Management Solutions, Chatsworth, Calif).

For orientation and reconstruction purposes, the skull was oriented in a sagittal plane in which the y-axis line was at the lower edge of the orbit and the x-axis was at the palatine bone. In the axial plane, the reconstruction was oriented such that the y-axis was placed at the midline of the skull and the x-axis was at the most basal point of the nasal cavity. In contrast, in the coronal plane, the y-axis was placed in the midline, with the x-axis being placed at the anterior nasal spine.

These patients had also completed a PSQ typically on the same day as the CBCT scan. The age of the children or adolescents ranged from 6-17 years with a mean age of 10.96 ± 2.61 years. There was no statistical significance in the body mass index between children/adolescents that who scored PSQ of <8 and PSQ of >8. Hence, we did not consider body mass index as a confounding factor. The inclusion criteria included all children and adolescents who had CBCT data of adequate image quality and completed PSQ records during the assessed period. The exclusion criteria included children and adolescents diagnosed with craniofacial anomaly syndrome. Children or adolescents with prior surgical intervention or trauma to their nasal region were also excluded. For simplicity, the children and adolescents will be referred to as patients throughout the manuscript. This specific sample should not be deemed representative of the general population as it involves patients seeking assessment for craniofacial problems and also patients that might be referred to a highly specialized sleep clinic assessing upper airway breathing problems.

The PSQ was developed on the basis of clinical experience, with concise and simple questions (responses: yes, no, or do not know). PSQ does not diagnose OSA; it is an accessory tool to identify patients at high risk. It can help prioritize patients requiring access to polysomnography (PSG), the reference standard for OSA, which does not have easy access and is expensive. PSQ has specific questions related to snoring, sleepiness, inattention or hyperactive behavior, and sleep-related breathing disorder. Moreover, the question items are grouped into 4 main domains: breathing, behavior, sleepiness, and other. Specifically, questions 1-8 and 14 belong to the breathing factor, questions 10 and 11 addressed sleepiness, questions 17-22 are related to the behavior factor, whereas the rest of the questions belonged to the other category. A positive PSQ was considered on the basis of ≥ 33% positive answers to the 22 validated question items (ie, a score of ≥ 8). No response questions or do not know questions were excluded from the calculation as suggested by Chervin et al.

Anatomic features of the nasal cavity were assessed in coronal and axial planes by 2 evaluators (C.T.B and P.B). Abnormalities were identified subjectively and objectively. The subjective assessment criteria were introduced as a way to represent a traditional clinical assessment of nasal septum deviation or TH.

For the subjective assessment of DNS, a nasal septum was subjectively considered deviated (DNS) when a slight to mild deviation was observed on the frontal view of the medial nasal septum of the CBCT scan ( Fig 1 ). The orientation used for the subjective assessment was the same for the objective assessment.

Fig 1
Subjective assessment of the nasal cavity anomalies. Two-dimensional CBCT scan images in the coronal and axial views were used to characterize a DNS and TH. Left panels : the nasal septum was considered deviated when a slight to mild deviation was observed ( circle ). Right panels : the turbinate was considered hypertrophic when 1 or both of the turbinates were enlarged ( arrow ).

For the objective assessment of DNS, 2 landmarks were identified in frontal view: (1) junction of the perpendicular plate with a cribriform plate of the ethmoid bone, and (2) junction of vomer bone with palatine bone. The digitalize/measure tool in the Dolphin Imaging software was used to measure the distance between the 2 landmarks, leading to an objective measure of the extent of septal deviation. The extent of deviation was calculated by the difference between the actual measured length of the deviated septum ( Fig 2 , a ) and the height of a hypothetical straight septum ( Fig 2 , b ), which was then divided by the height of the hypothetical straight septum. This value was converted into a percent. DNS was found to vary between 0% and 18%. A nasal septum (NS) was arbitrarily considered deviated when the value was ≥ 1.05, corresponding to at least a 5% deviation from the hypothetical straight line.

Fig 2
Objective assessment of the nasal cavity anomalies. Left panels : to calculate the degree of DNS, the length of DNS ( a ) in addition to the length of a hypothetical straight line ( b ) was calculated. Right panels : to measure the degree of TH, a ratio using the width of the turbinate itself ( c ) and the width of the space where the turbinate resides ( d ) was calculated. Coronal views were used to obtain the measurements.

For the subjective assessment of TH, turbinates were considered hypertrophic when they showed a visible size difference compared with the contralateral turbinate ( Fig 1 ). In 4 of the 99 patients, both-sided enlargement of turbinates was observed. Those patients were marked with turbinates fused with septum and considered hypertrophic.

For the objective assessment of TH, the ratio between the width of space occupied by turbinate soft tissue ( Fig 2 , c ) and the total width of the nasal cavity ( Fig 2 , d ) was determined and expressed in percentage (adapted from El-Anwar et al ). The turbinate was arbitrarily considered hypertrophic when the ratio was ≥80%, meaning ≥80% of the nasal cavity was obstructed by the turbinate. If both the turbinates were hypertrophic, the more severe hypertrophic turbinate was selected for the objective assessment.

To ensure reliability, all the 99 patients were assessed 2 times each by 2 evaluators (C.T.B and P.B), with each assessment being performed 3 days apart.

Reliability between 2 trained raters (C.T.B and P.B) was calculated using kappa for subjective assessments and intraclass correlation coefficient for objective assessments. In addition, descriptive, power, and effect size (partial eta-squared [η p 2 ]) correlation between variables was performed using chi-square tests and Pearson correlation coefficients. The level of significance was set at 0.05 in all statistical tests. The statistical procedure was carried out by SPSS (version 26.0; IBM, Armonk, NY) and JASP software for Windows.


Details on reliability assessment are shown in Supplementary Table I . Specifically, interrater reliability of subjective assessments of DNS was excellent, with kappa measurement of agreement between raters being 0.96. DNS objective assessment showed good intrareliability of 0.86 (95% CI, 0.55-0.95) and an interreliability of 0.79 (95% CI, 0.62-0.89). Interrater reliability of subjective assessment of TH was 0.93. In contrast, the objective assessment showed intrarater reliability of 0.89 (95% CI, 0.70-0.96) and interreliability of 0.88 (95% CI, 0.74-0.96) as shown in Supplementary Table I .

Of the 99 patients included in this study, 59 patients scored positive on the PSQ. Thus, these 59 patients were considered at high risk of SDB ( Table I ).

Table I
Characterization of sample demographics
Parameters n Sex Mean age, y Positive PSQ, >0.33
Total sample 99 57 male/42 female 10.96 (6-17) 59
Subjective 35 64 29 70
Objective (>5% DNS) (>80% TH) 72 27 74 25
Power 1.00 0.88

Subjective assessment of the nasal septum indicated that 64 of the 99 patients (64.65%) had DNS, whereas the objective assessment indicated that only 27 of the 99 patients (27.27%) had more than 5% septum deviation. Therefore, we decided on a 5% arbitrary cutoff to include most samples with a visual septum deviation. Anthropometric measurements of the nasal cavity showed that patients with shorter nasal cavity height (line A, 38.89 mm; 95% CI, 37.82-39.96 mm; P <0.009) and shorter NS length (line B 40.14 mm, 38.91-41.37 mm; P <0.029) presented significant more risk for SDB ( Supplementary Table II ). Overall, patients with a positive PSQ score showed a slightly higher mean for DNS (3.16%; 95% CI, 2.13-4.18) than patients with a negative PSQ score (2.74%; 95% CI, 1.48-4), but no statistically significant difference was observed, as seen in Supplementary Table II .

In regard to the subjective assessment of turbinates, 70 patients had some degree of TH (70.7%), whereas the objective assessment indicated that only 25 patients (25.3%) had a TH that occupied ≥80% nasal cavity space.

Males were the majority of participants in this study and presented the highest mean for PSQ score and TH. In addition, females presented the highest mean for DNS percentage. However, nonsignificant statistical differences were observed between the sexes. Additional descriptive statistics on sex, age, and objective and subjective nasal cavity assessments are outlined in Supplementary Tables II and III .

A statistical test of power was performed on the basis of the means and standard deviation. We established power of π = 0.8 as adequate for our tests; an appropriate power (π = 1.00) was observed for the nasal septum cluster analysis, both subjective and objective assessment, and TH ( Table I ).

To identify whether the nasal cavity abnormalities are associated with patients at high risk of SDB, a cross-tabulation statistical method between DNS, TH, and PSQ was used. Considering subjective assessments of DNS and TH, a statistically significant association ( P = 0.002) was identified between patients who were subjectively categorized to have both DNS and TH ( Table II ).

Table II
Clinical characteristics of children with DNS and TH
Parameters Subjective DNS Significance Objective DNS Significance
Non-DNS DNS P value Non-DNS DNS P value
PSQ code
Negative PSQ 42.9 39.1 0.713 44.4 29.6 0.181
Positive PSQ 57.1 60.9 55.6 70.4
Subjective TH
Non-TH 48.6 18.8 0.002 ∗∗ 33.3 18.5 0.149
TH 51.4 81.3 66.7 81.5
Objective TH (>80%)
Non-TH 74.3 25.0 0.938 75.0 74.1 0.925
TH 25.7 75.0 25.0 25.9

Oct 30, 2021 | Posted by in Orthodontics | Comments Off on Nasal cavity structural anomalies in children and adolescents at high risk of sleep-disordered breathing: An exploratory cone-beam computed tomography study
Premium Wordpress Themes by UFO Themes