A Study of the Facial Soft Tissue Morphology in Nasal- and Mouth-Breathing Patients

Abstract

Objectives

This study aimed to investigate the 3-dimensional (3D) facial morphology of children with skeletal Class II structure with different breathing patterns.

Methods

The 3dMDFace system (3dMD Inc.) was used to obtain 3D facial images. A total of 65 patients aged 10 to 12 years with skeletal Class II malocclusion (A point-Nasion-B point angle >5°) were grouped by sex into nasal-breathing (NB) and mouth-breathing (MB) participants. A total of 19 measurements, including linear distances, angles, and ratios, were measured. The measurements were compared using independent sample t test and Mann–Whitney U test. Factor analysis and logistic regression were used to test the correlation between facial morphology and different breathing patterns.

Results

For male children, the lower lip was longer in the MB group than in the NB group (P < .05). For female children, compared to NB, MB patients had a narrower mandibular width (P < .05), a smaller ratio of mandibular width to face height (MB: 0.99 ± 0.08 vs NB: 1.04 ± 0.09; P < .05), and a larger ratio of lower lip height to lip width (MB: 1.24 ± 0.10 vs NB: 1.19 ± 0.16; P < .05). In both male and female children, MB participants had a more convex nasolabial angle (P < .05) and an increased ratio of the lower part of the face to the upper facial height (male MB: 1.61 ± 0.17 vs male NB: 1.50 ± 0.12; female MB: 1.52 ± 0.10 vs female NB: 1.50 ± 0.20; P < .05). The logistic regression test showed no significant correlation between facial morphology and breathing patterns.

Conclusions

In participants with skeletal Class II pattern, MB children compared with NB children showed different facial morphology in the same sex group. The children with MB showed a more protruded upper lip and increased lower facial height, accounting for a larger proportion of the facial height. However, no significant correlation was found between facial morphology and breathing pattern. Only correlative trends were found.

Introduction

Mouth breathing is an undesirable habit that affects the muscular balance amongst the tongue, cheeks, and perioral muscles.¹ Chronic and allergic rhinitis have been associated with mouth breathing.² Inflammation or environmental irritation in the nasal cavity may lead to nasal mucosal edema, which decreases nasal ventilation and oxygen delivery efficiency.³

Hypertrophy of the adenoid and tonsil is considered the most prominent cause of mouth breathing due to the constriction of the respiratory tract, which could lead to the characteristic facial shape known as the adenoid face.^2,4 Generally, the adenoid grows rapidly after birth, reaches its maximum size in childhood, begins to regress at about 8 to 10 years old, and usually completely shrinks at 12 to 14 years old.⁵ On average, pubertal growth spurts occur at 12.0 to 14.4 years for boys and 9.8 to 11.5 years for girls.⁶ In other words, there is a temporal overlap between adenoid atrophy and this growth period. Although the adenoid shrinks at a certain time, the changes in facial shape remain for a lifetime. It is worth noting that adenoids may delay shrinkage under specified conditions.⁷ Excessive hypertrophy or delayed shrinkage of adenoids may lead to pharyngeal airway constriction, resulting in mouth breathing, especially during sleep.⁸

Adenoidectomy is the most common therapy for sleep-disordered breathing caused by adenoid hypertrophy.⁹ In recent years, the combined diagnosis of otolaryngology and orthodontics has become a recognised method for mouth-breathing diagnosis.¹⁰

Mouth breathing is a manifestation of oral dysfunction, which has a negative impact on the oral environment and craniofacial morphology.¹¹ It is reported that it may lead to excessive facial height, incompetent lips, narrow external nose, steep mandibular plane angle, “V” shaped maxillary arch, and retruded mandible.^9,12 Due to perioral muscle weakness and narrow maxillary arch, the upper anterior teeth and the lips protrude, further increasing the difficulty of lip closure.^2,11,13

Some studies have shown that the majority of mouth breathers have skeletal Class II malocclusion. Thus, this could be a risk factor for mouth breathing. The odds ratio was 1.73 as shown by Rossi et al.¹⁴ Also, a previous study has reported that the majority of patients with mouth breathing were characterised by retruded mandible.⁹

In recent years, 3-dimensional (3D) scanning technology has played a vital role in many fields. In orthodontics, it is mainly used for its high-precision facial measurements¹⁵ and is considered the gold standard for facial imaging,¹⁶ with high reliability and reproducibility.¹⁷ Nowadays, 3D scanning technology is trending in orthodontics, especially for facial analysis before and after orthodontic treatment. It has superiority over 2-dimensional photos due to its high capture speed, precision, and noninvasive method.¹³

Few studies about mouth-breathing patients were grouped according to age and sex.¹⁸ Moreover, most studies used cephalometric measurements during the growth and development period. This study aimed to evaluate the soft tissues of mouth-breathing and nasal-breathing children aged 10 to 12 years via 3D scanning technology.

Material and methods

Ethical approval

This retrospective study was approved by the Ethics Review Committee of Xi'an Jiaotong University with protocol number Xjkqll [2018] No.17. All patients' parents or guardians were informed and signed written consent forms to participate in the research.

Sample size calculation

The sample size was calculated according to the formula¹⁹:

$$n=f\left(\alpha ,\beta \right)\chi \frac{2{\sigma }^2}{{\left(\mu 1-\mu 2\right)}^2}$$

The sample size was calculated using a standard deviation from a previous study by Guimaraes et al.²⁰ to detect a difference of 5.26° in facial configuration angle (FCA) between mouth- and nasal-breathing patients.²⁰ The samples were further divided into male and female participants. At least 13 participants were required in each subgroup (α = 0.05, power of 80%).

Sample selection

All participants were patients in the Orthodontics Department of Xi'an Jiaotong University. The sample consisted of 65 participants (32 males and 33 females), all of whom were Chinese, including 35 mouth breathers and 30 nasal breathers (Supplementary Table 1).

Inclusion criteria included patients with skeletal Class II malocclusion, aged 10 to 12 years, with a normal body mass index (BMI) 13.43<BMI<21.68 kg/m².^21,22 Children who had a history of orthodontic treatment, cleft lip and palate, neurologic diseases, or facial defects and scars were excluded from this study.

Diagnosis of the breathing modes

An experienced orthodontist and an otolaryngologist assessed the patients’ breathing patterns. The orthodontist took a history from the patient's parents; they were asked about the existence of sleeping habits such as snoring, drooling on the pillow, or sleeping with an open mouth. Moreover, the Glatzel mirror test was performed to assist in the initial identification of mouth-breathing subjects.²³

Furthermore, the otolaryngologist performed a physical examination, including nasopharyngeal x-ray, rhinoscopy, and flexible nasopharyngoscopy. Participants who were found to have pharyngeal airway obstructions were diagnosed as mouth breathers.^24,25

Software and measurements

The 3dMDface system (3dMD Inc.) was used for data acquisition. Participants were instructed to look forwards and relax during scanning. The 3dMD programme combined the sequential face captures generated at a 1.5-millisecond capture speed, and the algorithms integrated the photos into a single 3D point cloud providing high-precision 3D images.

The data were measured by Geomagic Wrap 2017. To orient the scanned images, x-, y-, and z-axes were identified (x-axis: a plane connecting the right and left exocanthion points and parallel to the ground plane; y-axis: a plane passing through glabella point and perpendicular to the x-axis; z-axis: a plane passing through the top point of the right external auditory canal and perpendicular to the y-axis).

A single investigator who was blinded to the subjects’ demographic information was responsible for landmark identification and measurements^26,27 (Supplementary Table 2, Figure 1).

Fig. 1.

A, Facial soft tissue landmarks¹: glabella²; endocanthion_right³; endocanthion_left⁴; exocanthion_right⁵; exocanthion_left⁶; alae_right⁷; alae_left⁸; upper stomion⁹; lower stomion¹⁰; cheilion_right¹¹; cheilion_left¹²; pogonion¹³; mentolabial sulcus¹⁴; menton¹⁵; gonion_right¹⁶; gonion_left¹⁷; subnasale¹⁸; upper lip¹⁹; lower lip. B, Linear and angular measurements¹: upper facial height (UFH)²; upper lip height (ULH)³; lower lip height (LLH)⁴; facial configuration angle (FCA)⁵; nasolabial angle⁶; mentolabial sulcus angle.

Statistical analysis

The intraclass correlation coefficient (ICC) was used to assess the reliability of the measurements, and the standard error was calculated by the Dahlberg formula (SE = $\sqrt{\sum \frac{D^2}{2N}}$).²⁸

Chi-square test was used to detect the differences between sexes; Shapiro–Wilk test was performed to check the normal distribution of the measured variables; independent sample t test was used for normally distributed variables, and Mann–Whitney U test was used for the parameters that showed non-normal distribution.²⁹ Factor analysis was used to classify the variables; then, binary logistic regression was performed.

SPSS software (version 25.0) was used for data analysis, and the significance level was set at α = 0.05.

Results

Reliability of the measurements

The ICC was 0.93 to 0.99; the standard error was 0.5 to 0.7 mm for the linear measurements and 0.36° to 0.43° for the angular measurements.

Sex differences

The mandibular width (Go_R-Go_L), upper lip height (ULH), nasal width (Al_R-Al_L), and mentolabial sulcus angle were statistically significant between male and female participants. Boys presented with an increased mandibular width, nasal width (P < .01), and longer upper lip than girls (P < .01). The mentolabial sulcus angle was larger in girls (P < .05; Supplementary Figure 1; Supplementary Table 3).

Comparison between mouth-breathing and nasal-breathing patients

Mouth-breathing boys, compared to nasal-breathing boys, had a longer lower lip height (LLH; P < .05). (Table 1). On the other hand, mouth-breathing girls, compared to nasal-breathing girls, had a smaller mandibular width (P < .05); smaller ratio of the mandibular width to the sum of the upper facial height, upper lip height, and lower lip height (Go_R-Go_L/UFH+ULH+LLH; P < .05); and smaller ratio of the mandibular width to the sum of the height of the upper and lower lips (Go_R-Go_L/ULH+LLH; P < .05), whereas mouth-breathing participants in the female group showed a significantly increased value of lower lip height to lip width ratio (LLH/Ch_R-Ch_L; P < .05; Table 1).

Table 1.

Intragroup comparison between mouth-breathing and nasal-breathing patients in each sex separately.

			Men				Women
Measurements		MB, n = 17	NB, n = 15	P value		MB, n = 18 (mean ± SD)	NB, n = 15 (mean ± SD)	P value
Go_R-Go_L (mm)a		120.59 ± 8.33	119.07 ± 4.89	.271		110.85 ± 8.67	116.83 ± 8.22	.026*
En_R-En_L (mm)b		40.06 ± 2.91	40.56 ± 2.70	.273		38.85 ± 4.17	39.31 ± 2.47	.414
Ex_R-Ex_L (mm)a		87.70 ± 2.97	89.10 ± 4.30	.144		87.62 ± 5.18	86.69 ± 3.22	.274
Al_R-Al_L (mm)a		37.35 ± 3.11	36.67 ± 2.62	.254		34.96 ± 3.18	35.12 ± 1.90	.428
Ch_R-Ch_L (mm)a		39.35 ± 3.18	39.93 ± 3.15	.305		38.27 ± 3.27	39.26 ± 2.94	.185
UFH (mm)a		44.93 ± 3.96	45.35 ± 2.04	.356		44.73 ± 2.74	45.01 ± 2.67	.372
ULH (mm)a		22.72 ± 2.34	21.56 ± 2.14	.078		20.68 ± 1.74	20.89 ± 1.53	.364
LLH (mm)b		49.34 ± 4.82	46.51 ± 3.89	.037*		47.18 ± 3.59	46.43 ± 6.11	.091
FCA (°)a		18.06 ± 6.36	16.21 ± 4.11	.171		18.43 ± 6.23	18.16 ± 3.38	.439
Nasolabial angle (°)a		106.58 ± 9.71	113.07 ± 7.11	.021*		105.17 ± 7.38	112.37 ± 9.67	.011*
Mentolabial sulcus angle (°)b		140.82 ± 16.36	135.01 ± 18.89	.178		142.96 ± 17.31	147.73 ± 15.22	.282
Ratios
Go_R-Go_L/UFH+ULH+LLHa		1.03 ± 0.09	1.05 ± 0.05	.254		0.99 ± 0.08	1.04 ± 0.09	.027*
Go_R-Go_L/ULH+LLHa		1.68 ± 0.17	1.75 ± 0.10	.08		1.64 ± 0.14	1.74 ± 0.19	.032*
ULH+LLH/UFHb		1.61 ± 0.17	1.50 ± 0.12	.041*		1.52 ± 0.10	1.50 ± 0.20	.048*
Ch_R-Ch_L/Go_R-Go_La		0.33 ± 0.04	0.34 ± 0.03	.264		0.35 ± 0.03	0.34 ± 0.04	.186
Ch_R-Ch_L/Al_R-Al_La		1.06 ± 0.11	1.09 ± 0.07	.174		1.10 ± 0.11	1.12 ± 0.09	.300
Ch_R-Ch_L/Ex_R-Ex_La		0.45 ± 0.04	0.45 ± 0.02	.450		0.44 ± 0.04	0.45 ± 0.03	.108
ULH/LLHa		0.46 ± 0.06	0.47 ± 0.05	.475		0.44 ± 0.04	0.45 ± 0.05	.164
LLH/Ch_R-Ch_Lb		1.26 ± 0.18	1.17 ± 0.12	.067		1.24 ± 0.10	1.19 ± 0.16	.049*

**P < .01. *P < .05.

^aIndependent sample t test.

^bMann–Whitney U test.

In addition, the nasolabial angles were smaller in the mouth-breathing group than in the nasal-breathing group for both the male and female patients (P < .05), and mouth-breathing patients showed a greater ratio of the sum of upper and lower lip height to the upper facial height (ULH+LLH/UFH) than nasal-breathing patients as well (Table 1; Supplementary Figure 1).

Factor analysis

After repeated screening, the statistically significant variables were selected (Kaiser-Meyer-Olkin value > 0.5). The factor analysis classified the 8 variables into 3 dimensions (D1, D2, and D3), totally explaining 81.1% of the male and 82.9% of the female variables. D1 represented the smaller face height, explaining 39.4% of the male and 35.1% of the female variables; D2 represented the mandible width, explaining 26.8% of the male and 31.6% of the female variables; and D3 represented the lip protrusion, explaining 14.9% of the male and 16.2% of the female variables (Supplementary Table 4).

Binary logistic regression

The principal component scores of the 3 dimensions extracted from factor analysis were used to analyse their relationships with breathing patterns, and a binary logistic regression model was built (Table 2). The regression analysis did not show significant results (male group: D1, P = .05; D2, P = .906; D3, P = .893; female group: D1, P = .396; D2, P = .05; D3, P = .102). However, D1 for the male group (P = .05) and D2 for the female group (P = .05) tended to be significant. Based on this result, the receiver operating characteristic curve was drawn (Figure 2). Further results showed no significant relationship between facial measurements and breathing patterns for male patients. However, Go_R-Go_L/UFH+ULH+LLH and Go_R-Go_L/ULH+LLH showed a significant relationship with breathing patterns for female patients (Table 3).

Table 2.

Binary logistic regression model of 3 facial shape dimensions.

		B	SE	P value	OR	95% CI
						Lower	Upper
Male	D1 lower face height	0.823	0.419	.050	2.277	1.001	5.180
	D2 mandible width	−0.047	0.392	.906	0.954	0.442	2.060
	D3 lips protrusion	0.051	0.377	.893	1.052	0.503	2.201
	Constant	0.146	0.381	.702	1.157	—	—
Female	D1 lower face height	0.336	0.396	.396	1.399	0.644	3.039
	D2 mandible width	−0.954	0.488	.050	0.385	0.148	1.001
	D3 lips protrusion	−0.807	0.493	.102	0.446	0.170	1.172
	Constant	0.241	0.406	.553	1.272	—	—
Total	D1 lower face height	0.530	0.293	.070	1.699	0.957	3.018
	D2 mandible width	−0.488	0.280	.081	0.614	0.354	1.063
	D3 lips protrusion	−0.293	0.272	.280	0.746	0.438	1.270
	Constant	0.173	0.266	.514	1.189	—	—

Fig. 2.

Receiver operating characteristic (ROC) curve and diagnostic prediction of the breathing patterns. A, ROC curve for male patients. B, ROC curve for female patients. C, ROC curve regardless of sex.

Table 3.

Statistical analysis of receiver operating characteristic curve.

		AUC	P value	Sensitivity	Specificity	cutoff scores
Male	LLH	0.686	.073	0.588	0.8	49.5
	ULH+LLH/UFH	0.680	.082	0.529	0.867	1.62
	LLH/Ch_R-Ch_L	0.655	.136	0.412	1	1.305
Female	Go_R-Go_L	0.681	.076	0.444	0.933	108.8
	Go_R-Go_L/UFH+ULH+LLH	0.726	.027	0.889	0.667	1.035
	Go_R-Go_L/ULH+LLH	0.717	.034	0.889	0.667	1.745
Total	Factory scores	0.727	.002	0.468	0.967	0.670

AUC, area under the curve.

Discussion

The study demonstrated that male and female patients at similar ages with skeletal Class II malocclusion presented different facial soft tissue characteristics in the following variables: GO_R_GO_L, Al_R-Al_L, ULH, and mentolabial sulcus angle. Furthermore, intragroup comparison of breathing modes showed significant differences between mouth-breathing and nasal-breathing patients in these parameters: LLH, nasolabial angle, and ULH+LLH/UFH for boys. Moreover, females showed significance in Go_R-Go_L, nasolabial angle, Go_R-Go_L/UFH+ULH+LLH, Go_R-Go_L/ULH+LLH, ULH+LLH/UFH, and LLH/Ch_R-Ch_L.

The children in this study had normal maxillary growth, and the results showed no significant difference in FCA angle between mouth- and nasal-breathing patients; this was consistent with studies by Souki et al³⁰ and Jakobsone et al.³¹ However, it contradicted the findings of Inada et al11 for preschool children, in which they did not consider sagittal classifications.

In the current study, boys had significantly longer upper lip height than girls. However, intragroup comparison between mouth-breathing and nasal-breathing groups showed no significant difference. Shimizu et al³² claimed that mouth breathers had longer upper lip than nasal breathers, but their study did not consider sex-related differences. Our findings suggested that sex differences might have an impact on the upper lip height.

The mouth-breathing boys with skeletal Class II malocclusion presented a significantly longer lower lip than the nasal-breathing boys with skeletal Class II malocclusion. This suggests that the retruded mandible might be due not only to the underdevelopment of the mandible but also to the downwards and backwards rotation of the mandible, which is known as the clockwise rotation of the mandible. This might place the mandible into a retruded position in relation to the cranial base; several studies have shown that mouth-breathing patients’ clockwise rotation of the mandible might be related to a disproportionate increased anterior lower vertical face height in relation to posterior facial height.^1,2,18,33,34

In the female group, the mandibular width (Go_R-Go_L) was smaller in mouth-breathing than nasal-breathing patients, which implied that the lower facial part was narrower. Numerous studies demonstrated that the lower position of the tongue in the mouth-breathing patient reduces the tongue pressure on the maxilla,³⁵ which will lead to inadequate transverse development of the maxilla.³³ In order to adapt to the occlusal function, a compensatory lateral hypoplasia may occur in the mandible.³⁶ The masticatory muscles of male patients might have greater masticatory strength than those of female patients.³⁷ Thus, the stronger masticatory muscles might conceal the lack of lateral development of the mandible, which might explain the lack of difference in mandibular width between mouth-breathing and nasal-breathing participants in the male group. Previous studies have found that decreased mandibular width was associated with constricted pharyngeal airway.^38,39 Consequently, reduced pharyngeal airway was linked to mouth breathing.⁴⁰ Thus, mandibular hypoplasia could be associated with mouth breathing, but it is not necessary a causal relationship.

In this study, Go_R-Go_L/UFH+ULH+LLH and Go_R-Go_L/ULH+LLH were both smaller in mouth-breathing girls than their nasal-breathing counterparts. However, regarding vertical facial parameters, UFH, ULH, and LLH in the female group did not show significant differences between mouth and nasal breathing. This means that mouth breathing probably affects the proportions of the lateral to the vertical facial development, leading to a long facial appearance.

The ratio of lower lip height to lip width was greater only in mouth-breathing girls than in nasal-breathing girls. This finding suggested that mouth-breathing girls appeared to have relatively small mouths; a previous study found that sex dimorphism was associated with differences in lip morphology.⁴¹

In this study, the upper facial height showed no statistical difference, whilst the ratio of the sum of the upper and lower lips to the upper face height was greater in the mouth-breathing group, suggesting that mouth-breathing participants tended to have a more elongated lower face than nasal-breathing participants; this was consistent with previous studies.^2,9 A previous study reported that the ratio of lower anterior facial height to the total anterior facial height was greater in mouth-breathing patients.⁴² However, those measurements were based on hard tissue points from the lateral cephalograms.

In addition, the nasolabial angle in the mouth-breathing group was smaller than that in the nasal-breathing group (P < .05). This implies that the upper lip was more protruded in the mouth-breathing group, which might be related to the constriction of maxillary arch in mouth-breathing patients.^11,30

Previous research found that the lower lip of patients with mouth breathing was shorter than that in the nasal-breathing group.³⁰ However, in this study, the lower lip of mouth-breathing boys was longer than that of nasal-breathing boys This contradiction might be related to the different facial 3D measurement technique, sex grouping, landmarks, and parameters used in this study.

The logistic regression was used to verify the correlation between the 3D soft tissues and the breathing patterns. No significant correlation was found. However, the P values of D1 (lower face height) in the male group and D2 (mandible width) in the female group were close to .05. It speculated that the increase of lower face height and the decrease of mandibular width tended to be related to the occurrence of mouth breathing.

There are still some disputes about whether to collect longitudinal or cross-sectional data for children in their development period. Longitudinal research can verify the changes in an individual's craniofacial complex in different development stages, but sometimes there are some difficulties in data collection, which makes it challenging to track and follow up the research with a large sample size regularly in the experimental stage. On the contrary, cross-sectional studies have various advantages; the most important is that a large number of samples spanning different periods and their various growth stages can be studied in the same group of investigators simultaneously and the cost can be controlled.⁴³ This study adopted a cross-sectional design on the premise of combining the above advantages. However, the cross-sectional design is not capable of assessing the causality principle. Thus the findings of this study must be interpreted with caution. The age range was reduced to 10 to 12 years to reduce the error caused by the differences amongst individuals of a wide range of ages.

To the best of our knowledge, almost all previous literature compared mouth- and nasal-breathing participants as if the sex difference would not be a confounding factor. In the current study, sex differences in facial soft tissue were assessed. Significant differences were found between male and female participants. Thus, participants in each breathing pattern were matched by sex to reduce the error caused by sex and to assess the effect of breathing mode on the facial soft tissue more accurately .

This study only studied skeletal Class II malocclusion, the most common sagittal classification for mouth-breathing patients. Due to the lack of samples, we did not include the vertical skeletal patterns in this study so that the results could be affected by the different vertical growth patterns of the participants. Also, the study did not include skeletal Class I and skeletal Class III. More detailed grouping would be necessary for further in-depth research.

Conclusions

The following conclusions can be drawn based on the results of the current study. Sex-related discrepancy in the mandibular width, inter-alar distance, upper lip height, and mentolabial sulcus angle was observed. Mouth breathing affected the lower third of the face. Male mouth-breathing patients showed longer facial soft tissues than their nasal-breathing counterparts, whilst female mouth breathing compared to female nasal breathing resulted in narrower mandibular width. Moreover, the nasolabial angle of mouth breathing patients was smaller than nasal breathers of both sexes. The increase in lower facial height and the decrease in mandibular width seemed to be related to mouth breathing.

Author contributions

Bo Cheng: methodology, data curation, conceptualisation, and software; Amin S. Mohamed: writing original draft, investigation, and software; Janvier Habumugisha: formal analysis, visualisation, and resources; Yucheng Guo: draft review and editing; Rui Zou: conceptualisation; Fei Wang: supervision, methodology, draft review and editing, and project administration. All authors reviewed the results and approved the final version of the manuscript.

Funding

• 1.General project from the field of social Development, in Department of Science and Technology of Shaanxi Province, Grant/Award Number: 2019SF-081
• 2.Science and Technology Project of Xi 'an, Grant/Award Number: 20YXYJ0010(3)
• 3.Clinical New Technology from Stomatological Hospital of Xi'an Jiaotong University in 2018

No funders had any involvement in the study design, data collection, analysis, interpretation, writing of the report, and decision to submit the article for publication.

Conflict of interest

None disclosed.