Mechanical Constriction of the Maxilla Alters Nasal Architecture

Abstract

Introduction: We investigated the effect of transverse maxillary constriction on nasal septal deviation (NSD) and nasal floor slanting. Methods: 60 growing Wistar rats (21 days old) were divided into four groups: (1) Experimental Group 1 received active constriction force (100cN), (2) Experimental Group 2 received active expansion force (100cN), (3) Sham received the same spring as Experimental Groups without receiving any active force, and (4) Control group did not receive any appliance. Samples were collected after 28 days for microcomputed tomography (μCT) analysis. Results: Experimental Group 1 demonstrated maxillary constriction (both skeletal and dental), accompanied by mandibular shift on closure, clockwise mandibular rotation, and increased mandibular plane angle and facial height. Constriction was also associated with severe nasal floor slanting in the molar area that extended posteriorly. Nasal floor canting was accompanied by a slanted vomer and a C-shaped NSD. The direction of nasal floor canting and mandibular shift was always similar. Experimental Group 2, on the other hand, was not associated with nasal deviation, and a slight slanting of the nasal floor was observed only when there was a mandibular shift. Conclusions: Our study suggests that the constricting transverse forces applied to the maxilla can be associated with nasal septal deviation. One possible mechanism by which constriction contributes to nasal septal deviation is by promoting mandibular shift. Mandibular shift, in turn, dictates the direction of slanting of the nasal floor and, consequently, the vomer, which may, in turn, lead to nasal septal deviation.

1. Introduction

Why we have a nasal septal deviation is not clear. Acquired and congenital factors that contribute to the development of nasal septal deviation, such as trauma and craniofacial anomalies, are well-studied [1,2,3,4,5,6,7,8,9,10]. However, these factors alone cannot explain the high incidence of nasal septal deviations. This is important because nasal septal deviation is among the leading causes of nasal obstruction, along with enlarged adenoids, concha bullosa, and inferior turbinate hypertrophy, all of which may lead to mouth breathing. Chronic mouth breathing over an extended period can increase the constricting forces on the maxilla, thereby affecting its transverse and vertical growth [11,12,13]. Further maxillary constriction can cause occlusal disharmony, narrowing of the pharyngeal airway, alterations in tongue posture resulting in retroglossal airway narrowing, and potentially contribute to obstructive sleep apnea [14,15,16,17,18,19].

Previous studies have shown that nasal septal deviation is rare among long-snouted mammals and is more common in mammals with shorter faces [3,20]. Similarly, from an evolutionary point of view, it has been suggested that humans may have developed a high incidence of nasal septal deviation as a result of the continued diminution of our face [20,21]. It has also been reported that individuals with a more retrognathic nasomaxillary complex exhibit a greater magnitude of septal deviation compared to individuals with a more mesognathic midface [9]. In addition, nasal septal deviation is more common in individuals with reduced facial height [22]. Similarly, experimental reduction in the maxillary length through surgery restricts normal septal growth and induces septal deviation [23]. These observations suggest a possible developmental etiology for nasal septal deviation, possibly due to a constricted maxilla. Based on this theory, another factor that may cause deviations is when cartilage volume exceeds the available space for the cartilage due to a decrease in the maxilla’s transverse dimension. This suggests reciprocal effects between the nasal septum and maxilla; a constricted maxilla produces nasal septal deviation, and in turn, nasal septal deviation through mouth breathing worsens the maxillary constriction.

However, studies investigating maxillary transverse deficiency as a developmental factor in nasal septal deviation are rare and indirect, focusing mainly on associations between nasal septal deviation and maxillary constriction [24,25,26,27].

Here, we further investigate the effect of maxillary constriction on the development of nasal septum deviation in growing rats, and the mechanism by which maxillary constriction may contribute to nasal septal deviation. Considering that the incidence of maxillary constriction is increasing in humans [3,28,29,30,31,32,33,34,35,36,37], establishing a relationship between maxillary constriction and nasal septal deviation development may have significant health implications.

2. Materials and Methods

2.1. Animal Study

Growing Wistar rats (n = 60, average weight 47 g, 21 days old) were studied at two centers: Constriction experiments were completed at Universidad Nacional Autónoma de México (UNAM), while the Expansion experiments were completed at New York University, All experiments were approved by the Institutional Animal Care and Research Advisory Committee (CICUAL, ID 10389, approved 9 August 2024), from Universidad Nacional Autónoma de México (UNAM), and the National Research Council’s Guide for the Care and Use of Laboratory Animals from New York University (IACUC Protocol # 151003, approved 9 September 2018). Although using 2 centers was not ideal, to avoid introducing variability, the same investigator fabricated, calibrated, and delivered the constriction and expansion appliances at both centers. Collected specimens were fixed and then transferred to the same facility for analysis consistency. Therefore, all specimen analyses were conducted at the same imaging facility, using the same microCT equipment (µCT; Skyscan1172; Bruker microCT, Kontich, Belgium), and measurements were completed by the same investigators.

Rats at this stage are in a growing phase, and their molars are fully erupted. All animals were housed in polycarbonate cages in a 12 h light/dark environment at a controlled temperature of 21–23 °C and relative humidity of 40–60%. Animals were housed in an individually ventilated rack system and fed a specially prepared soft diet with tap water ad libitum. The mixture consisted of 200 g of standard chow (Teklad, Envigo, Indianapolis, Indiana, IN, USA), 600 mL of water, and 45 g of gelatin, which was homogenized and subsequently molded into bars to facilitate consumption by the rats. This diet was used to ensure adequate nutritional intake while minimizing the possibility of dislodging the appliance.

For randomization, the animals were numbered 1 to 60, and 15 random numbers were subsequently generated for each group, using the standard = RAND() function in Microsoft Excel. The four groups were: (1) an untreated Control, (2) a Sham Group, (3) Experimental Group 1 (Constriction), and (4) Experimental Group 2 (Expansion). Allocation was not concealed, as the effects of constriction or expansion could be observed during daily assessment of appliance integrity.

2.2. Transverse Force Application

On day 0, animals from the Sham and Experimental Groups were sedated with isoflurane delivered through the SomnoSuite anesthesia system (Kent Scientific, Torrington, CT, USA). Isoflurane was administered at 1% during the initial sedation phase, increased to 3% to achieve deep anesthesia, and subsequently reduced to 0.5% for maintenance. Anesthesia was verified by lack of response to toe-pinch. Springs were fabricated from 0.016″ stainless steel wires (3M Unitek, Monrovia, CA, USA). The Experimental Groups received a calibrated, custom-designed constriction or expansion spring that delivered compression or tensile transverse force (100cN) to the molars (Figure 1). This force was selected based on previous studies demonstrating that high cellular activity was observed in the suture at 100cN [38]. Considering that rats would have a normal vertical chewing force, averaging between 54 and 75 N, this force is not regarded as excessive [39]. The springs were fabricated, calibrated, and delivered by the same investigator at both centers. While rats were sedated, the springs were cemented with composite resin on the interproximal and occlusal surfaces of the molars by the same investigator. The Sham group received a similar spring, but it was not activated and produced no forces. Compressive or tensile forces were applied for 28 days. Animals were monitored daily under inhalational anesthesia with isoflurane–nitrous oxide as the standard method of general anesthesia to ensure the integrity of the appliance. No appliances were dislodged or needed to be reinstalled. Therefore, all animals were included in the analysis and final results.

2.3. Micro-CT Imaging

Animals were euthanized by CO₂ narcosis on day 28, and samples were collected for micro-CT analysis (μCT). After euthanasia, the whole skull was dissected and fixed for over 72 h with 4% (w/v) paraformaldehyde in 0.1 M phosphate buffer, pH 7.4, followed by storage in 70% ethanol.

The whole maxilla was scanned by micro-computed tomography (µCT; Skyscan1172; Bruker microCT, Kontich, Belgium). The specimens were scanned at 13.55 µm voxel size, 100 KV, 0.300 degrees rotation step (192.30 degrees angular range), and a 1910 ms exposure per view in 70% ETOH. Results were analyzed utilizing the NRecon (1.6.9.16) software on the HP open platform (OpenVMS Alpha Version 1.3-1 session manager) for 3D reconstruction and viewing of images. Superimpositions were performed using Amira 6.0.0.

2.4. Intra-Oral and Intra-Nasal Measurements

Palatal width (at the level of intersection between alveolar bone and palatal walls) and interdental width (the distance between the height of contour of the first molars) were analyzed in the mid-coronal plane at the level of the upper first molar from the three-dimensional images of µCT (Figure 2A). Changes in the upper molar inclination were studied by measuring the angle (a) formed by the long axes of the upper right and left first molars in the same plane.

The slant of the nasal floor was measured on the coronal plane of the µCT image, specifically in the area where the slant was most pronounced. One line was drawn tangent to the nasal floor and parallel to the orbits, and another line was drawn along the steepest slant of the nasal floor. The angle between these lines was measured and recorded as the slant of the nasal floor (Figure 2B). On the same µCT image, the palatal height was measured as the perpendicular distance from the highest point on the palate to the occlusal plane (Figure 2B).

The degree of the nasal septum deviation was assessed by measuring the angle between the tangent line to the upper part of the nasal septum and a perpendicular line drawn from the horizontal line connecting the center of the two orbits. (Figure 2C). The width of the left and right nasal cavities was measured on the area of maximum deviation on a horizontal line from the septum to the lateral walls of the nasal cavity (Figure 2C). Measurements were performed on soft tissue view of the coronal section of microCT images, where the deviation was most significant. Investigators completing these morphometric measurements were blinded to group allocation to avoid bias.

2.5. Extra-Oral Measurements

Skull changes in response to constriction were studied by measuring the angle between the mandibular plane (the line tangent to the lower border of the mandible) and the palatal plane (the line tangent to the palate). The anterior facial height was measured on the perpendicular line to the mandibular plane from the most anterior point of the nasal bone (Figure 3A). The posterior facial height was measured from the highest point in the condylar process to the lowest point on the gonial angle (Figure 3B). Maximum condylar width was measured in the coronal view from the medial to the lateral heights of contour (Figure 3C).

2.6. Statistical Analysis

Two examiners completed all morphological quantifications. The random and systematic errors were calculated using the formula described by Dahlberg and Houston [40,41]. The random and systematic errors were found to be small both intra-observer (0.015 and 0.017 mm, respectively) and inter-observer (0.018 and 0.02 mm, respectively).

Significant differences between the Experimental Groups, Shams, and Controls were assessed using analysis of variance (ANOVA). A pairwise multiple-comparison analysis was performed using Tukey’s post hoc test. Two-tailed p-values were calculated; p < 0.05 was set as the level of statistical significance.

Table 1. Transverse changes in the maxilla in response to constriction.

	Intermolar Width	Palatal Width	Intermolar Angle
Control	4.8 ± 0.2 mm	2.8 ± 0.1 mm	45.8 ± 0.8°
Sham	4.9 ± 0.2 mm	2.8 ± 0.1 mm	46.2 ± 0.9°
Constriction	3.6 ± 0.3 mm *	1.9 ± 0.3 mm *	24.3 ± 1.4° *

Intermolar width and palatal width at the level of mid-first molar. Data represent mean ± SEM of 15 samples (* significantly different from Control; p < 0.01 by ANOVA test). All three comparisons remained significant after Holm–Bonferroni correction (adjusted p < 0.0001).

,

Table 2. General skull characteristics after 28 days of constriction.

	Angle Between the Mandibular Plane and the Palatal Plane	Anterior Facial Height	Posterior Facial Height	Condylar Width
Sham	8° ± 1°	15.7 ± 0.8 mm	12.4 ± 0.7 mm	1.81 ± 0.2 mm
Constriction	15° ± 2° *	20.2 ± 1.3 mm *	9.2 ± 0.8 mm *	1.23 ± 0.3 mm *

Data represent mean ± SEM of 15 samples (* significantly different from Sham, p < 0.01 by ANOVA test). All four comparisons remained significant after Holm–Bonferroni correction (adjusted p < 0.0001).

,

Table 3. Nasal floor inclination

	Nasal Floor Angle	Palatal Depth
Sham	0 ± 1°	3.1 ± 0.1 mm
Constriction	26 ± 3° *	4.8 ± 0.5 mm *

Data represent mean ± SEM of 15 samples (* significantly different from Sham, p < 0.01 by ANOVA test). Both comparisons remained significant after Holm–Bonferroni correction (adjusted p < 0.0001).

and

Table 4. Nasal septal deviation.

	Nasal Septal Angle	% Difference in the Width of the Left and Right Nasal Cavities
Sham	0 ± 2°	0 ± 0.1%
Constriction	11 ± 3° *	16 ± 2% *

Data represent mean ± SEM of 15 samples (* significantly different from Sham, p < 0.01 by ANOVA test). Both comparisons remained significant after Holm–Bonferroni correction (adjusted p < 0.0001).

report 11 continuous morphometric outcomes across four anatomical domains (dental arch, Table 1; craniofacial skeleton, Table 2; nasal floor, Table 3; nasal septum, Table 4), measured on an overlapping set of specimens. To control the family-wise error rate across this full set of comparisons, a Holm–Bonferroni step-down correction was applied to the 11 outcome-level tests (Constriction vs. Control and sham for Table 1; Constriction vs. Sham for Table 2, Table 3 and Table 4), with α = 0.05. Holm’s method was chosen over a fixed Bonferroni correction for its greater statistical power, and over a false-discovery-rate approach because these are pre-specified, confirmatory comparisons rather than an exploratory screen.

Posterior crossbite, anterior crossbite, and mandibular shift (

Table 5. Association between maxillary constriction and occlusal characteristics.

Characteristic	Constriction (n = 15)	Sham (n = 15)	p-Value
Posterior crossbite	15 (0.6 + 0.1 mm)	0 (0 mm)	1.29 × 10−8
Anterior crossbite	15 (1.5 + 0.5 mm)	0 (0 mm)	1.29 × 10−8
Mandibular shift	15 (1.2 + 0.5 mm)	0 (0 mm)	1.29 × 10−8

Data represent the number of animals per group (n = 15/group); parenthetical values are the mean + SEM of the linear displacement among affected animals, averaged across 2 independent examiners. Presence was defined as displacement exceeding the linear measurement error established for this dataset (intra-observer 0.015–0.017 mm; inter-observer 0.018–0.02 mm; Dahlberg-Houston method). Fisher’s exact test, two-tailed, was applied to each 2 × 2 contingency table; all 3 comparisons remain significant after Holm–Bonferroni correction for m = 3 (most conservative threshold, α/3 = 0.0167). Inter-examiner agreement on categorical classification was perfect (Cohen’s κ = 0.91).

) were treated as present when the measured displacement exceeded the intra- and inter-observer measurement error established above (≥0.1 mm for posterior crossbite; ≥0.5 mm for anterior crossbite and mandibular shift), and compared between groups with Fisher’s exact test (two-tailed). Two examiners independently classified each animal, and inter-examiner agreement was assessed using Cohen’s κ (0.91). A Holm–Bonferroni correction was applied across this family of 3 comparisons (α = 0.05).

Directional laterality of mandibular shift, nasal floor slanting, and nasal septal deviation within the Constriction group (

Table 6. Association between mandibular shift and skull characteristics.

Finding	Constriction	Sham	p-Value
A. Own-direction distribution vs. chance (exact binomial, two-tailed, p0 = 0.50)
Mandibular shift (left vs. right)	9/6	—	0.607
Nasal floor slanting (left vs. right)	9/6	—	0.607
Nasal septal deviation (left vs. right)	8/7	—	1.000
B. Directional concordance within the constriction group (Exact Binomial Test vs. p0 = 0.50)
Mandibular shift direction = nasal floor slanting direction	15/15	—	6.1 × 10−5
Mandibular shift direction = nasal septal deviation direction	14/15	—	9.8 × 10−4
Direction of palatal slant = direction of nasal floor slant	15/15	—	6.1 × 10−5

The data represent animals in the Constriction and Sham groups. Direction (left/right) was assigned only when the underlying signed measurement exceeded the operational cutoff defined in Methods (nasal septal deviation ≥ 5°; nasal floor slanting ≥ 5°); mandibular shift direction used the ≥0.5 mm threshold established in Table 5. Sham animals exhibited none of these phenomena (Table 5) and are therefore not a comparator here; Table 6 addresses the directional pattern only among Constriction animals (n = 15). All 6 comparisons (Parts A and B) form a single Holm–Bonferroni family (family-wise α = 0.05; step-down thresholds 0.0083, 0.010, 0.0125, 0.0167, 0.025, 0.050). Ranked by ascending raw p-value, the three concordance comparisons (Part B) clear their respective thresholds and remain significant; the three own-direction comparisons (Part A) do not depart significantly from chance (raw p ≥ 0.607), consistent with the non-significant directional trend already reported in Results Section 3.2, Section 3.3 and Section 3.4.

) used the same mandibular shift threshold (≥0.5 mm) and an angular threshold of ≥5° for nasal floor slanting and nasal septal deviation, chosen because it falls within the non-overlapping range separating Sham from Constriction on these measures (Table 3 and Table 4). For each phenomenon, the observed left/right split was tested against a 50:50 chance expectation with an exact binomial test (two-tailed). The same binomial approach was used to test cross-phenomenon concordance—whether the direction of one phenomenon matched another within the same animal—against a null concordance of 0.50. Sham animals were excluded from these analyses, since none showed a shift, slant, or deviation in either direction (Table 5). The resulting six comparisons (three own-direction tests, and three concordance tests) were treated as one Holm–Bonferroni family (α = 0.05).

A multivariate model (MANOVA or ANCOVA) was considered for both outcome families but not pursued. With n = 15/group, the sample size does not support stable multivariate covariance estimation for the largest domain in either family (4 correlated continuous outcomes, Table 2; 6 categorical/directional outcomes, Table 6), and the distributional assumptions these models require could not be confidently verified at this sample size. More importantly, every univariate comparison in this study showed a large effect (p ≤ 0.017 for all comparisons; p < 0.0001 for all continuous outcomes and 5 of 6 categorical outcomes), so a multivariate approach would not alter any conclusions. Holm–Bonferroni correction of the univariate tests was therefore retained as the more conservative and defensible approach throughout.

3. Results

3.1. Constriction Significantly Decreased the Palatal Width, Interdental Width, and Dental Angulation

Application of constricting force to the maxilla was accompanied by narrowing of the upper dental arch in the area of the molars (Figure 4A). To better understand the nature of this narrowing, the width of the palate, interdental width, and angulation between the long axis of the teeth were analyzed at 28 days at the level of the mid-coronal plane of the upper first molar from the three-dimensional images of µCT as described (Figure 2A).

Constriction significantly decreased dental and palatal width compared with the Control and Sham groups at 28 days, with a statistically significant difference (p < 0.01) (Table 1). The palatal movement of the teeth and alveolar bone includes significant up-righting of molars as observed with a substantial decrease in the angulation between the long axis of the upper first molars, which was statistically significant (p < 0.01). The magnitude of dental changes was greater than that of skeletal changes, as demonstrated by a greater decrease in dental width than in palatal width. Since no difference was observed between the Sham and Control groups, only data from the Sham group is presented in the rest of the article.

Maxillary constriction caused the animals to shift their mandibles into a comfort zone for biting, and, in these experiments, nine of the fifteen rats in the constriction group moved to the left side; however, the difference in the direction of shift between animals was not statistically significant (p > 0.05). The result of the mandibular shift was an asymmetric bite in the anterior teeth (Figure 4B), leading to changes in the mesio-distal angulation of some of the incisors (Figure 4C). This also produced a posterior crossbite (Figure 4D) with a consequent adaptive change in the inclination of the lower molars, with more inclination of the posterior molars on the non-crossbite side.

3.2. Constriction of the Maxilla Resulted in Clockwise Rotation of the Mandible

Constriction of the upper jaw and narrowing of the dental arch were accompanied by relative extrusion of maxillary teeth (up-righting of molars in the coronal plane), which caused clockwise rotation of the mandible and an increase in the mandibular plane angle compared to the palatal plane. This was accompanied by an increase in anterior facial height and a decrease in posterior facial height, both of which were statistically significant (p < 0.01) (Figure 5A,B) (Table 2). An increase in mandibular plane angle was accompanied by an increase in overjet (the horizontal relation between the upper and lower incisors) and a decrease in overbite (the vertical relation between the upper and lower anterior teeth), resulting in an animal with a Class II appearance (Figure 5A). Change in biomechanics of the jaw was accompanied by condylar adaptation and a decrease in the width of the condyles that was statistically significant (p < 0.01) (Figure 5C).

3.3. Maxillary Constriction Was Associated with Nasal Floor Slanting

µCT images of Experimental Group 1 rats revealed an average slant of 26° in the nasal floor, which is statistically significant compared to the Sham Group (p < 0.01). None of the Sham or Control Groups demonstrated any nasal floor deformity (Table 3).

In Experimental Group 1, nine animals demonstrated a slant extending from right to left. In comparison, in six animals, the slant was extended from left to right with no statistical difference (p > 0.05) (Figure 6A). The direction of slant and the direction of the shift in the mandible were the same.

The deformity of the nasal floor started in the area of the molars, where the constricting force was applied, but not in the anterior region of the septum. The slant in the nasal floor was continuous, varying in degree, and extended to the posterior nasal cavity (Figure 6B).

The palate followed the nasal floor slanting. The direction of the palatal slant in all animals was the same as that of the nasal floor slant. In addition, a significant increase in palatal depth was observed compared with the Sham group (Table 3).

The nasal floor and palatal floor deformities also affected the suture morphology (Figure 6C). The deformity did not change the horizontal location of the mid-palatal suture; however, it changed the coronal direction of the suture, which demonstrated that the coronal direction of the mid-palatal suture can be altered due to constricting forces. The direction of the sutural tilt was independent of the direction of slanting (Figure 6C).

3.4. Maxillary Constriction Was Associated with Nasal Septal Deviation

Maxillary constriction was accompanied by curvature of the nasal septum. Nasal septum deviation was limited to the area of the molars; the anterior part of the nasal septum did not demonstrate any deviation. While the most frequent septal direction of deviation was toward the lowest side of the nasal floor slant, fewer animals demonstrated opposite deviation with no statistical difference (p > 0.05) (Figure 7A). The magnitude of deviation from the vertical position in the average was approximately 11°, which was significant compared to the Sham Group (p < 0.01) (Table 4). In our experiments, the deviation was always C-shaped.

While in the sham group, the left and right nasal cavities were similar in size, in the experimental group, the size of the nasal cavities was significantly wider on the opposite side of the nasal septal deviation (Table 4) (p < 0.03).

The vomer bone, which holds the nasal septum in a channel structure, demonstrated a tilt in the channel. This tilt followed the nasal septum deviation and was more frequent toward the lower part of the nasal septal deviation (Figure 7B).

The potential relationship between maxillary constriction and occlusion characteristics is summarized in Table 5.

The potential relation between mandibular shift, a consequence of maxillary constriction, and different skull characteristics has been summarized in Table 6.

3.5. Mandibular Shift Contributed to the Nasal Floor Slanting

To determine whether nasal floor slanting was related to the mandibular shift, we reversed the direction of the force and applied tensile force (expansion force). In response to tensile forces, expansion of the dental arches was observed. While some animals did not show a shift in the mandible, some shifted to the left, and some shifted to the right, with no significant difference in trend (p > 0.05) (

Table 7. Nasal floor slanting in response to expansion.

Expansion Group	Number of Animals	Nasal Floor Slanting
No mandibular shift	6	0 ± 2
Mandibular shift to the left	4	7 ± 3 *
Mandibular shift to the right	5	8 ± 3 *

All the nasal floor slants were in the same direction as the direction of mandibular shift. Data represent mean ± SEM of 15 samples (* significantly different from Sham, p < 0.05).

). The animals that did not show a mandibular shift did not show slanting of the nasal floor (Figure 8A), while the animals that received tensile forces and developed a mandibular shift developed a mild slant. The direction of the slant and mandibular shift was the same (Figure 8B). This observation demonstrated that mandibular shift can contribute to nasal floor slanting in both constriction and expansion groups. However, in the expansion group, no septal deviation was observed.

4. Discussion

Various factors can contribute to nasal septal deviation, including increased intrauterine pressure, trauma during birth, trauma from injuries, genetic factors, and nasal floor asymmetry [1,2,3,4,5,6,7,8,9,10]. Here, we explore whether constriction of the maxilla can also contribute to the development of nasal septal deviation.

Our experiments demonstrate that maxilla constriction affects nasal septum shape. However, it does not establish if this effect is direct or indirect. This agrees with previous findings that found a correlation between maxillary width and nasal septal deviation in humans [42,43]. It should be emphasized that this observation does not claim that maxillary constriction is the only etiological factor in nasal septal deviation, but it increases clinicians’ awareness of the need to explore maxillary constriction as a contributing factor.

The causes of the constriction can be due to a number of environmental factors, including ecological allergens/nasal obstruction, dietary changes, and decreased rates of breastfeeding [28,29,30,31,32,33,34,35,36,37]. However, how the maxillary constriction contributes to the development of nasal septum deviation is not clear. Three possibilities exist.

Hypothesis 1: One possible way maxillary constriction may affect the growth of the nasal septum is by creating a growth mismatch between the septum and maxilla. This mismatch may cause the nasal septum (cartilage and bone) to bulge away from the midsagittal plane. This hypothesis aligns with previous studies in humans, which have shown that when normal midline nasal septal growth is disrupted, the frequency of septal deviation increases [3,9,20,44,45].

For this hypothesis to be correct, we need to demonstrate that the nasal septum exhibits robust interstitial growth that persists even in the absence of the space required for growth of the septum due to maxillary constriction. But which part of the nasal septum has the interstitial growth?

The nasal septum is composed of three parts: the cartilaginous septum and the bony septum, which is composed of the perpendicular plate of ethmoid and the vomer. The ossification of the vomer via the intramembranous pathway typically is completed by birth, and the ossification of the perpendicular plate accelerates until age 10, then slows significantly afterward [46,47]. It has been established that the bony part of the septum does not have interstitial growth and primarily functions as a growth site [48,49]. On the other hand, since the nasal cartilage is considered the anterior extension of the chondrocranium, it is regarded as a primary cartilage and, therefore, has interstitial growth. It has been suggested that this interstitial growth can produce enough mechanical forces that may help in the maxilla’s displacement and suture activation and, therefore, cause maxillary growth [50,51]. Thus, the nasal septum has been proposed as the growth center for the maxillary complex [8,52,53,54,55,56,57,58,59]. While this claim is evident primarily in long-snouted animal models [60], it is controversial for the human facial skeleton [61]. One group believes that the nasal septum in humans is similar to that of many long-snouted animals and acts as a growth center for the maxilla [6,7,8,9,53,57,61], and nasal septal deviation can produce a force that can cause asymmetry in the skull [61,62,63,64,65]. On the other hand, many others could not find the evidence that the mechanical force produced by nasal septum interstitial growth was strong enough to displace the maxilla. These scientists believe that the soft tissue (periosteal matrix) and oronasal spaces (capsular matrix) are responsible for maxillary growth [48,66,67], and nasal septum growth is secondary to midfacial growth. They believe that the nasal septum is more important for transferring occlusal forces to the skull, rather than for the growth of the maxillary complex [61,68,69]. The opponents to the role of the nasal septum as a growth center argue that if the mechanical force produced by the nasal septum were large enough, it would prevent the development of nasal septum deviation in the first place by pushing the maxilla downward and forward.

While the interstitial growth of the nasal septum is undeniable, the magnitude of its contribution to nasal septal deviation is questionable. This is because the majority of the nasal septum’s growth decreases significantly by the age of two, while maxillary growth continues by age 14–16 [70]. Based on this observation, nasal septal growth can only partially explain the development of septal deviation.

Hypothesis 2: The second possibility that may explain why maxillary constriction affects the nasal septum deviation is that constriction pushes the palatal shelves upward toward the space occupied by the nasal septum, causing a deviation in the septum. This bulging in response to constriction can be explained mechanically by the appearance of significant moments in the system (Figure 9). We argue that the constriction forces applied far from the center of resistance of each hemimaxilla can produce considerable moments in the system that push the maxilla up in the midline toward the nasal cavity and decrease the space for the septum (Figure 9B). If this hypothesis is correct, one would expect the application of tensile forces to produce opposite moments and decrease bulging in the middle of the nasal floor, as observed in our experiment (Figure 10). This analysis is in agreement with previous observations that nasal septal deviation is associated with an increase in palatal arch height [3,4,11,71]. Our current data does not examine this hypothesis since, in our model, maxillary constriction was always associated with mandibular shift. However, this hypothesis remains a strong possibility biomechanically.

Symmetrical moments produced by constricting forces should cause bulging in the middle of the nasal floor, but no slanting should occur. In other words, mechanical constriction cannot explain the nasal septal floor slanting, since it produces symmetrical forces and moments. However, in our experiment, a significant nasal floor slant was observed, indicating that mechanical factors were unequal and disrupted the balance between the right and left nasal floors. Therefore, constricting forces alone cannot explain the slant in the nasal floor, suggesting the possibility of other factor(s). Here, we propose that the mandibular shift, occurring in response to the maxillary constriction, can play a role in the development of nasal floor slanting. It should be emphasized that mandibular shift is always secondary to maxillary constriction or expansion, as we observed here, and does not occur separately, especially in animals in the absence of any trauma.

Maxillary constriction prevents proper contact between the posterior teeth and forces the mandible to shift to one side to establish a posterior occlusion, resulting in a crossbite, as observed in our experiments. This produces large horizontal forces toward the medial side of the non-crossbite side. These forces produce an additional moment that can push the palate even further up on that side (Figure 11). On the other hand, on the crossbite side, the horizontal forces remain small and directed laterally, producing a lateral moment that rotates the palate in the opposite direction of the moment created by constricting forces, thereby reducing the magnitude of the rotation of the palatal shelf on that side. This differential moment between the left and right contributes to the slanting of the nasal floor.

In our study, no specific trend was observed in the shift in the mandible toward one side. However, the direction of shift in the mandible and nasal floor cant was always the same. To further test the hypothesis that mandibular shift contributes to the slanting of the nasal floor, we studied animals subjected to expansion (tensile forces). Here, we expected no nasal floor slanting. However, similar in animals subjected to constriction forces, the bite in these animals changed, and the majority of the animals demonstrated a shift in their mandibles toward a normal bite on one side and a scissor bite on the other (Figure 8). However, some animals were able to continue biting in the center. In none of the animals did the expander cause bulging of the nasal floor as was expected. In the absence of mandibular shift, no slanting of the nasal floor was observed. But in the presence of mandibular shift, mild to moderate nasal floor slanting was observed. In this scenario, similar to a constriction group, the direction of nasal floor slanting and mandibular shift was the same. These experiments together demonstrate that the slanting was caused by mandibular shift and could be exaggerated by constricting forces. It should be emphasized that the shift itself did not cause the large horizontal forces and moments. The shift is associated with a change in the biomechanics of occlusion, such as a change in the inclination of the lower molars, that will continue even as the mandible gradually shifts, leading to adaptation of form and the development of permanent asymmetry. It should also be noted that since the changes in nasal septal deviation occur throughout life [9,46,72,73], one can expect the dynamic changes in occlusion that can occur throughout life to contribute to changes in nasal deviation.

In our experiment, we did not establish whether the effect of constriction on nasal septal deviation is direct (Hypothesis 1 or 2) or indirect through mandibular shift (Hypothesis 3). This is due to a study limitation, which shows that constriction is always associated with mandibular shift, and separating the effect of one from the other was not possible. However, we are currently developing a model to induce mandibular shift in the absence of constriction or expansion to study the isolated effect of mandibular shift on nasal cavity architecture. Based on current data, we proposed the third hypothesis, that constriction-induced mandibular shift may indirectly affect nasal septal deviation. We know that the nasal floor is part of the palatal process of the maxilla, and its growth is a combination of bony resorption on the nasal side, along with bony apposition over the oral surface, which is considered part of natural cortical drift [49]. The change in loading of the maxilla and the appearance of moments and horizontal forces due to constriction forces and mandibular shift produce cortical drifting in the opposite direction, which causes changes in the floor of the nose and the height of the palate [74,75]. We believe that this change in occlusal forces and the appearance of new moments that caused nasal floor slanting can be the third source for the development of nasal deviation. Nasal floor slanting is accompanied by a change in the direction of the vomer, as was observed in these experiments. This is important, since the vomer holds the base of the nasal septal cartilage (Figure 12). Nasal septal cartilage lies in the vomerine groove, a shallow channel at the superior border of the vomer bone. Therefore, a change in the direction of the vomer could easily cause nasal septal deviation. Our observation that nasal floor slanting and vomer inclination could be considered as a third factor in the development of nasal septal deviation is in agreement with previous studies, which show a high incidence of nasal septal deviation associated with nasal floor slanting [65,70,76].

It should be emphasized that hypothesis three that mandibular shift can cause nasal floor slanting (always in the same direction) and may affect nasal septal deviation by changing the direction of vomer is in agreement with the observation that in our experiment all mandibular shifts were associated with change in direction of vomer. However, while vomer and nasal septal deviation were always in the same direction, the nasal floor slanting and the vomer inclination were not always in the same direction. This differs from previous studies, which noted that nasal septum deviation is more pronounced on the lower side of the nasal floor [3,43,76,77]. This difference may be due to the limited sample size in our experiment. These observations suggest that there is a complex mechanical environment in the oral and nasal cavities, and that we are far from understanding the full aspect of these interactions.

In our experiments, constriction of the maxilla was accompanied by C-shape deformation of the nasal septum, in accordance with the classification system proposed by Guyuron [78]. However, we did not observe any other type of deformation. This may be due to a short study duration. Our observation is in agreement with previous reports that have shown the most frequent nasal septal deviation associated with nasal floor slanting is the C-shape deformity [43]. Perhaps that can explain why the application of expansion forces can be more successful in the treatment of C-shape deviation, as has been reported before [79].

This study sheds light on nasal floor slanting as the third source of nasal septal deviation. However, in addition to nasal septal deviation, nasal slanting has other side effects. Studies demonstrated that severe slanting of the nasal floor may also contribute to nasal obstruction [80], which emphasizes the importance of clinical diagnosis of constricted maxilla that can cause mandibular shift and indirectly contribute to nasal floor slanting.

Rats in this experiment demonstrated clockwise rotation of the mandible, increased lower facial height, and decreased posterior facial height. These changes were not due to a change in the animal’s breathing pattern and were induced mostly by constriction of the upper arch and uprighting of the upper posterior teeth. However, they produced similar skeletal effects to those observed in mouth-breathing patients [81,82]. This similarity can be due to the effect that mouth breathing has on the constriction of the upper arch [71]. Chronic mouth breathing can increase the constricting forces on the maxilla, which will affect the transverse and vertical growth of the maxilla [11,12,13].

It should be emphasized that our study neither tested nor denied the possible effect of septal growth on the maxilla [51,54,55,56]. However, we demonstrated that even in the presence of the mechanical forces from the septum, maxillary growth can be substantially altered by numerous other factors. In fact, the maxilla itself can change the direction of septal growth, which can cause chronic blockage of the nasal cavity and mouth breathing [36,83,84,85]. Based on this reciprocal theory, deformity of craniofacial structures worsens nasal obstruction and chronic mouth breathing, which, in turn, leads to craniofacial maldevelopment. This negative feedback loop highlights the need to interrupt the cycle as early as possible through proper diagnosis and early interceptive treatment. While current data indicate a potential negative feedback loop between the oral and nasal cavities, these findings should not be over-interpreted or generalized to humans without further validation. Further research in this area, especially in humans, is required.

5. Conclusions

This article contributes to our understanding of the interaction between the oral and nasal cavities. First, it sheds light on maxillary constriction as a possible developmental factor in nasal septal deviation. Second, this study emphasizes that the nasal cavity and oral cavity are two systems that have a significant reciprocal effect on each other’s form. Third, it identified two possible pathways by which maxillary constriction can affect nasal septal deviation: on one pathway, maxillary constriction creates mechanical forces and moments that cause bulging of nasal floor, which invades the space for the septum; and on the other pathway, maxillary constriction can change occlusal forces, which may lead to mandibular shift which in turn may affect nasal floor slanting, vomer deviation, and contribute to nasal septal deviation. Finally, this article confirms the clinical observation that maxillary constriction affects the overall skeletal form and produces deformations. Given the use of a growing rat model with artificial orthodontic forces and the lack of long-term follow-up, these findings are best viewed as hypothesis-generating; if ultimately supported by longer-term human studies, they would strengthen the rationale for early diagnosis and interceptive treatment of maxillary constriction. Children and pre-pubertal individuals show greater adaptability of the nasal septum following orthopedic interventions, such as maxillary expansion [86]. This is in addition to the other beneficial effects of palatal expansion on increasing the nasal cavity and decreasing inspiratory nasal resistance [87,88]. Based on this study, there is a significant need for additional research in this field and for closer interdisciplinary collaboration to understand the side effects of rising incidents of maxillary constriction [3,28,29,30,31,32,33,34,35,36,37].